import pandas as pd
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
import matplotlib.gridspec as gridspec
import mpl_toolkits as mplot3d
import scipy as sp
pd.set_option("display.max_columns", None)
from tqdm import tqdm
tqdm.pandas()

def lighten_color(color, amount=0.5):
    """
    Lightens the given color by multiplying (1-luminosity) by the given amount.
    Input can be matplotlib color string, hex string, or RGB tuple.

    Examples:
    >> lighten_color('g', 0.3)
    >> lighten_color('#F034A3', 0.6)
    >> lighten_color((.3,.55,.1), 0.5)
    """
    import matplotlib.colors as mc
    import colorsys
    try:
        c = mc.cnames[color]
    except:
        c = color
    c = colorsys.rgb_to_hls(*mc.to_rgb(c))
    return colorsys.hls_to_rgb(c[0], 1 - amount * (1 - c[1]), c[2])

DATA CLEANING AND WRANGLING

# DF1 = Total and facet scores, and score per question
df0 = pd.read_csv('Indian Risk Survey_Besample_Filtered_12_2024.csv')
df1 = df0.iloc[:, 19:62].copy()
df1.insert(0, "ResponseId",0)
df1["ResponseId"] = df0["ResponseId"].copy()
df1

	ResponseId	Q0	Q1	Q2	Q3	Q4	Q5_1	Q6_1	Q8_1	Q8_2	Q8_3	Q8_4	Q8_5	Q8_6	Q8_7	Q8_8	Q9b_1	Q9b_4	Q9b_5	Q9b_6	Q9b_7	Q9b_8	Q9b_9	Q11a_1	Q11a_2	Q11a_3	Q11b	Q11c	Q11d	Q11h	Attention Check	Q11i	Q12a	Q12b	Q12c	Q13a	Q13b	Q13c	Q13d	Q14a	Q14b	Q14c	Q15a	Q14b.1
0	R_4RadSQzbD3kAfEP	Yes, I would like to participate in the study ...	Male	18	Secondary School (11th to 12th Std.)	Student	4	7	3	3	3	4	5	4	3	4	25.0	15.0	10	10	10.0	10	20.0	2	2	2	HIV test,Professional horoscope	Irrespective of the quality of the study, futu...	The growth rate over five years will be betwee...	5 in 100 people prior to the intervention to 6...	Vase	The medication increases recovery by 100%	More than $102	Less than today with the money in this account	False	500	10	0.10	9 out of 10	25.0	20	They are equal	INR 1500,001 – INR 30,00,000	10000000.0
1	R_4F8RqnI7xnwXmBY	Yes, I would like to participate in the study ...	Male	41	Undergraduate Program	Salaried/Employee/Consultant in a sector other...	6	9	4	4	4	4	4	5	4	5	10.0	30.0	0	0	20.0	40	0.0	2	1	1	HIV test,Fingerprint,DNA test	Irrespective of the quality of the study, futu...	It is not possible to predict the growth rate ...	100 in 10,000 people prior to the intervention...	Vase	The medication increases recovery by 100%	More than $102	Less than today with the money in this account	False	500	10	0.10	59 out of 1000	25.0	20	Can’t say	INR 500,001 – INR 15,00,000	5000000.0
2	R_8taN3wUPbeZMxzM	Yes, I would like to participate in the study ...	Male	20	Undergraduate Program	Student	3	0 (no investment experience)	6 (strongly agree)	6 (strongly agree)	4	1 (strongly disagree)	1 (strongly disagree)	2	1 (strongly disagree)	6 (strongly agree)	30.0	40.0	10	10	0.0	5	5.0	1	1	1	Fingerprint	The higher the quality of the study, the more ...	It is not possible to predict the growth rate ...	70 in 100 people prior to the intervention to ...	Vase	The medication increases recovery by 50%	More than $102	More than today with the money in this account	Do not know	1000	10	50.00	9 out of 59	50.0	20	They are equal	< INR 500,000	10000.0
3	R_4KT4eUqEbo2q5s9	Yes, I would like to participate in the study ...	Male	18	Secondary School (11th to 12th Std.)	Student	5	0 (no investment experience)	6 (strongly agree)	6 (strongly agree)	4	4	1 (strongly disagree)	6 (strongly agree)	4	6 (strongly agree)	0.0	0.0	0	20	40.0	40	0.0	1	1	2	Fingerprint,DNA test	The higher the quality of the study, the more ...	The growth rate over five years will be exactl...	100 in 10,000 people prior to the intervention...	Vase	The medication increases recovery by 50%	More than $102	Less than today with the money in this account	False	500	10	0.10	59 out of 1000	25.0	20	They are equal	< INR 500,000	1000000.0
4	R_4TMr0yMiNpsJBr5	Yes, I would like to participate in the study ...	Male	37	Undergraduate Program	Salaried/Employee/Consultant in a sector other...	3	1	4	5	5	6 (strongly agree)	6 (strongly agree)	6 (strongly agree)	6 (strongly agree)	2	70.0	10.0	0	0	0.0	10	10.0	2	1	1	HIV test,Fingerprint,DNA test,Cancer screening...	The higher the quality of the study, the more ...	The growth rate over five years will be exactl...	100 in 10,000 people prior to the intervention...	Vase	The medication increases recovery by 100%	More than $102	Less than today with the money in this account	False	500	10	0.10	59 out of 1000	25.0	20	They are equal	< INR 500,000	1000000.0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
365	R_47eRbAoGWc5Ttzb	Yes, I would like to participate in the study ...	Male	48	Post-Graduate Program	Salaried/Employee/Consultant in a sector other...	0 (unwilling to take risk)	1	4	5	4	5	5	2	3	6 (strongly agree)	50.0	10.0	10	20	2.0	5	3.0	1	1	1	Professional horoscope	Irrespective of the quality of the study, futu...	The growth rate over five years will be exactl...	70 in 100 people prior to the intervention to ...	Vase	The medication increases recovery by 2%	Refuse to answer	Exactly the same as today with the money in th...	False	500	68	0.01	59 out of 100	50.0	40	They are equal	< INR 500,000	50000.0
366	R_4eWLoP4wkX5K1kl	Yes, I would like to participate in the study ...	Male	22	Undergraduate Program	Salaried/Employee/Consultant in a sector other...	5	6	5	5	4	5	5	6 (strongly agree)	5	4	70.0	15.0	0	0	15.0	0	0.0	2	1	2	Cancer screening test	The higher the quality of the study, the more ...	The growth rate over five years will be exactl...	It is not possible to determine which of the a...	Vase	The medication increases recovery by 2%	More than $102	More than today with the money in this account	False	500	50	0.10	9 out of 59	25.0	20	They are equal	< INR 500,000	800000.0
367	R_47rbcPZByLBrZFn	Yes, I would like to participate in the study ...	Male	19	Diploma and Vocational Training	Student	5	5	6 (strongly agree)	6 (strongly agree)	6 (strongly agree)	6 (strongly agree)	6 (strongly agree)	5	6 (strongly agree)	5	10.0	20.0	10	15	25.0	15	5.0	1	2	1	DNA test	Irrespective of the quality of the study, futu...	It is not possible to predict the growth rate ...	It is not possible to determine which of the a...	Vase	The medication increases recovery by 2%	More than $102	Less than today with the money in this account	False	345	1	0.10	9 out of 10	90.0	20	They are equal	INR 500,001 – INR 15,00,000	2500000.0
368	R_42y9IJJWALsuHPF	Yes, I would like to participate in the study ...	Male	28	Undergraduate Program	Salaried/Employee/Consultant in a sector other...	6	7	6 (strongly agree)	3	5	5	6 (strongly agree)	6 (strongly agree)	6 (strongly agree)	5	4.0	0.0	0	0	6.0	40	50.0	2	1	1	HIV test,Fingerprint,DNA test,Cancer screening...	The higher the quality of the study, the more ...	The growth rate over five years will be betwee...	70 in 100 people prior to the intervention to ...	Vase	The medication increases recovery by 100%	More than $102	Less than today with the money in this account	False	600	10	0.10	9 out of 10	25.0	5	Hertinol	INR 1500,001 – INR 30,00,000	4500000.0
369	R_8k6D0jzzHCC5X3Z	Yes, I would like to participate in the study ...	Male	25	Undergraduate Program	Salaried/Employee/Consultant in a sector other...	7 (willing to take risk)	8	6 (strongly agree)	5	5	4	6 (strongly agree)	6 (strongly agree)	6 (strongly agree)	5	5.0	5.0	5	5	30.0	20	30.0	2	1	1	DNA test	The higher the quality of the study, the more ...	It is not possible to predict the growth rate ...	5 in 100 people prior to the intervention to 6...	Vase	The medication increases recovery by 100%	More than $102	Less than today with the money in this account	False	500	500	50.00	59 out of 1000	25.0	10	Hertinol	> INR 75,00,000	8000000.0

370 rows × 44 columns

Demographics

df1["age"] = df1["Q2"]

# Age groups/rec = [2,3,4]
# Age group 2 = 18 to 35 y/o
# Age group 3 = 36 to 59 y/o
# Age group 4 = 60 to 75 y/o (75 y/o, i.e, within the scope of the data we have, it can mean 60 and above also)

df1["age_rec"] = np.where( (df1["age"] >= 16) & (df1["age"] <= 35), 2,
                          np.where( (df1["age"] >= 36) & (df1["age"] <= 59), 3,
                                   np.where( (df1["age"] >= 60), 4, 0
                                           )
                                  )
                         )


# Secondary upto 10 – ISCED 3
# Senior Secondary upto 12 - ISCED 3 
# Diploma and voca – ISCED 2
# UG – ISCED 2
# PG – ISCED 1
# PhD and higher – ISCED 1

df1["isced"] = np.where(df1["Q3"] == "Post-Graduate Program", 1,
                        np.where(df1["Q3"] == "Undergraduate Program", 2,
                                 np.where(df1["Q3"] == "Ph.D. and higher", 1,
                                          np.where(df1["Q3"] == "Diploma and Vocational Training", 2,
                                                   np.where(df1["Q3"] == "Secondary School (11th to 12th Std.)", 3,
                                                            np.where( df1["Q3"] == "Primary School (up to 10th Std.)", 3,
                                                                     np.where( df1["Q3"] == "M.Phil.", 1, 0
                                                                             )
                                                                    )
                                                           )
                                                  )
                                         )
                                )
                       )

df1["income"] = np.where(df1["Q15a"] == "< INR 500,000", 1,
                         np.where(df1["Q15a"] == "INR 500,001 – INR 15,00,000", 2,
                                  np.where(df1["Q15a"] == "INR 1500,001 – INR 30,00,000", 3,
                                           np.where(df1["Q15a"] == "INR 30,00,001 – INR 50,00,000", 4,
                                                    np.where(df1["Q15a"] == "INR 50,00,001 – INR 75,00,000", 5,
                                                             np.where( df1["Q15a"] == "> INR 75,00,000", 6,0
                                                                     )
                                                            )
                                                   )
                                           )
                                  )
                         )

df1["wealth"] = df1["Q14b.1"]

Questions

matchCol = ["q8_2_1", "q8_2_2", "q8_2_3", "q8_2_4", "q8_2_5", "q8_3", "q8_4", "q8_5berlin_1", 
            "q8_5london_1", "q8_5paris_1", "q8_6", "q8_7", "q9_1_1", "q9_2_1", "q9_3",
            "q10_1_1", "q10_2_1", "q10_3_1", "q10_4"]
df1[ matchCol ] = 0

df1.columns

Index(['ResponseId', 'Q0', 'Q1', 'Q2', 'Q3', 'Q4', 'Q5_1', 'Q6_1', 'Q8_1',
       'Q8_2', 'Q8_3', 'Q8_4', 'Q8_5', 'Q8_6', 'Q8_7', 'Q8_8', 'Q9b_1',
       'Q9b_4', 'Q9b_5', 'Q9b_6', 'Q9b_7', 'Q9b_8', 'Q9b_9', 'Q11a_1',
       'Q11a_2', 'Q11a_3', 'Q11b', 'Q11c', 'Q11d', 'Q11h', 'Attention Check',
       'Q11i', 'Q12a', 'Q12b', 'Q12c', 'Q13a', 'Q13b', 'Q13c', 'Q13d', 'Q14a',
       'Q14b', 'Q14c', 'Q15a', 'Q14b.1', 'age', 'age_rec', 'isced', 'income',
       'wealth', 'q8_2_1', 'q8_2_2', 'q8_2_3', 'q8_2_4', 'q8_2_5', 'q8_3',
       'q8_4', 'q8_5berlin_1', 'q8_5london_1', 'q8_5paris_1', 'q8_6', 'q8_7',
       'q9_1_1', 'q9_2_1', 'q9_3', 'q10_1_1', 'q10_2_1', 'q10_3_1', 'q10_4'],
      dtype='object')

df1["q8_2_1"] = np.where(df1["Q11b"].str.contains("HIV test"), 1, 2)
df1["q8_2_2"] = np.where(df1["Q11b"].str.contains("Fingerprint"), 1, 2)
df1["q8_2_3"] = np.where(df1["Q11b"].str.contains("DNA test"), 1, 2)
df1["q8_2_4"] = np.where(df1["Q11b"].str.contains("Cancer screening test"), 1, 2)
df1["q8_2_5"] = np.where(df1["Q11b"].str.contains("Professional horoscope"), 1, 2)

df1["q8_3"] = np.where(df1["Q11c"] == "The lower the quality of the study, the more likely that future studies will change the risk estimate.", 1,
                          np.where(df1["Q11c"] == "The higher the quality of the study, the more likely that future studies will change the risk estimate.", 2,
                                  np.where(df1["Q11c"] == "Irrespective of the quality of the study, future studies will not change the risk estimate.", 3,
                                          np.where(df1["Q11c"] == "Irrespective of the quality of the study, future studies will change the risk estimate substantially anyway.", 4,0
                                                  )
                                          )
                                  )
                         )


df1["q8_4"] = np.where(df1["Q11d"] == "The growth rate will be 0.4% on average each year", 1,
                          np.where(df1["Q11d"] == "The growth rate over five years will be exactly 2%", 2,
                                  np.where(df1["Q11d"] == "The growth rate over five years will be between 1% and 3%", 3,
                                          np.where(df1["Q11d"] == "It is not possible to predict the growth rate with certainty", 4,0
                                                  )
                                          )
                                  )
                         )

df1["q8_5berlin_1"] = np.where(df1["Q11a_1"] == 1, 1, 2)
df1["q8_5london_1"] = np.where(df1["Q11a_1"] == 1, 1, 2)
df1["q8_5paris_1"] = np.where(df1["Q11a_1"] == 1, 1, 2)

df1["q8_6"] = np.where(df1["Q11i"] == "The medication increases recovery by 100%", 1,
                          np.where(df1["Q11i"] == "The medication increases recovery by 50%", 2,
                                   np.where(df1["Q11i"] == "The medication increases recovery by 2%", 3,
                                            np.where(df1["Q11i"] == "None of the above is implied", 4, 0
                                                    )
                                           )
                                  )
                         )

df1["q8_7"] = np.where(df1["Q11h"] == "5 in 100 people prior to the intervention to 6 out of 100 people after the intervention", 1,
                          np.where(df1["Q11h"] == "100 in 10,000 people prior to the intervention to 120 out of 10,000 people after the intervention", 2,
                                   np.where(df1["Q11h"] == "70 in 100 people prior to the intervention to 90 out of 100 people after the intervention", 3,
                                            np.where(df1["Q11h"] == "It is not possible to determine which of the answers is correct given the information provided", 4, 0
                                                    )
                                           )
                                  )
                         )


df1["q9_1_1"] = df1["Q14a"].copy()
df1["q9_2_1"] = df1["Q14b"].copy()

df1["q9_3"] = np.where(df1["Q14c"] == "Crosicol", 1,
                          np.where(df1["Q14c"] == "Hertinol", 2,
                                   np.where(df1["Q14c"] == "They are equal", 3,
                                            np.where(df1["Q14c"] == "Can’t say", 4, 0
                                                    )
                                           )
                                  )
                         )

df1["q10_1_1"] = df1["Q13a"].copy()
df1["q10_2_1"] = df1["Q13b"].copy()
df1["q10_3_1"] = df1["Q13c"].copy()

df1["q10_4"] = np.where(df1["Q13d"] == "9 out of 59", 1,
                          np.where(df1["Q13d"] == "9 out of 10", 2,
                                   np.where(df1["Q13d"] == "59 out of 1000", 3,
                                            np.where(df1["Q13d"] == "59 out of 100", 4, 0
                                                    )
                                           )
                                  )
                         )

df1["ResponseId"].nunique()

370

scoreColumns = ["certainty1", "certainty2", "certainty3", "certainty4", "certainty5", "uncertainty1", "uncertainty2", "numeracy1", "numeracy2", "numeracy3", "numeracy4", "numeracy5", "graph1", "graph2", "graph3", "riskcalculation1", "riskcalculation2", "riskcalculation3", "riskcalculation4"]
df1[scoreColumns] = 0

# Assigning scores

def scoring1(df1Facet):
    df1Facet.loc[ df1Facet["q8_2_1"] == 2, "certainty1"] = 1
    df1Facet.loc[ df1Facet["q8_2_2"] == 2, "certainty2"] = 1
    df1Facet.loc[ df1Facet["q8_2_3"] == 2, "certainty3"] = 1
    df1Facet.loc[ df1Facet["q8_2_4"] == 2, "certainty4"] = 1
    df1Facet.loc[ df1Facet["q8_2_5"] == 2, "certainty5"] = 1

    df1Facet.loc[ df1Facet["q8_3"] == 1, "uncertainty1"] = 1
    df1Facet.loc[ df1Facet["q8_4"] == 4, "uncertainty2"] = 1

    df1Facet.loc[ df1Facet["q8_5berlin_1"] == 2, "numeracy1"] = 1
    df1Facet.loc[ df1Facet["q8_5london_1"] == 1, "numeracy2"] = 1
    df1Facet.loc[ df1Facet["q8_5paris_1"] == 1, "numeracy3"] = 1
    df1Facet.loc[ df1Facet["q8_6"] == 1, "numeracy4"] = 1
    df1Facet.loc[ df1Facet["q8_7"] == 4, "numeracy5"] = 1

    df1Facet.loc[ df1Facet["q9_1_1"] == 25, "graph1"] = 1
    df1Facet.loc[ df1Facet["q9_2_1"] == 20, "graph2"] = 1
    df1Facet.loc[ df1Facet["q9_3"] == 3, "graph3"] = 1

    df1Facet.loc[ df1Facet["q10_1_1"] == 500, "riskcalculation1"] = 1
    df1Facet.loc[ df1Facet["q10_2_1"] == 10, "riskcalculation2"] = 1
    df1Facet.loc[ (df1Facet["q10_3_1"] == 0.1) | (df1Facet["q10_3_1"] == ".1") | (df1Facet["q10_3_1"] == ",1") , "riskcalculation3"] = 1
    df1Facet.loc[ df1Facet["q10_4"] == 1, "riskcalculation4"] = 1

    # Assigning total scores
    #df1Facet["Certainty score_5"] = df1Facet["certainty1"] + df1Facet["certainty2"] + df1Facet["certainty3"] + df1Facet["certainty4"] + df1Facet["certainty5"]
    #df1Facet["Uncertainty score_5"] = df1Facet["uncertainty1"] + df1Facet["uncertainty2"]

    #df1Facet["Number Comprehension score_5"] = df1Facet["numeracy1"] + df1Facet["numeracy2"] + df1Facet["numeracy3"] + df1Facet["numeracy4"] + df1Facet["numeracy5"]
    #df1Facet["Graph Comprehension score_5"] = df1Facet["graph1"] + df1Facet["graph2"] + df1Facet["graph3"]

    #df1Facet["Calculation score_4"] = df1Facet["riskcalculation1"] + df1Facet["riskcalculation2"] + df1Facet["riskcalculation3"] + df1Facet["riskcalculation4"]

    #df1Facet["Total Score_19"] = df1Facet["Certainty score_5"] + df1Facet["Uncertainty score_5"] + df1Facet["Number Comprehension score_5"] + df1Facet["Graph Comprehension score_5"] + df1Facet["Calculation score_4"]

    return df1Facet

df1Facet = df1.groupby(["ResponseId"]).progress_apply(scoring1)

100%|███████████████████████████████████████████████████████████████████████████████| 370/370 [00:02<00:00, 172.64it/s]

df1Facet

		ResponseId	Q0	Q1	Q2	Q3	Q4	Q5_1	Q6_1	Q8_1	Q8_2	Q8_3	Q8_4	Q8_5	Q8_6	Q8_7	Q8_8	Q9b_1	Q9b_4	Q9b_5	Q9b_6	Q9b_7	Q9b_8	Q9b_9	Q11a_1	Q11a_2	Q11a_3	Q11b	Q11c	Q11d	Q11h	Attention Check	Q11i	Q12a	Q12b	Q12c	Q13a	Q13b	Q13c	Q13d	Q14a	Q14b	Q14c	Q15a	Q14b.1	age	age_rec	isced	income	wealth	q8_2_1	q8_2_2	q8_2_3	q8_2_4	q8_2_5	q8_3	q8_4	q8_5berlin_1	q8_5london_1	q8_5paris_1	q8_6	q8_7	q9_1_1	q9_2_1	q9_3	q10_1_1	q10_2_1	q10_3_1	q10_4	certainty1	certainty2	certainty3	certainty4	certainty5	uncertainty1	uncertainty2	numeracy1	numeracy2	numeracy3	numeracy4	numeracy5	graph1	graph2	graph3	riskcalculation1	riskcalculation2	riskcalculation3	riskcalculation4
ResponseId
R_1YkMM2lMB9aEuVL	150	R_1YkMM2lMB9aEuVL	Yes, I would like to participate in the study ...	Female	41	Undergraduate Program	Salaried/Employee/Consultant in a sector other...	3	3	3	4	3	3	3	3	3	4	85.0	0.0	0	0	5.0	10	0.0	1	1	1	DNA test	The higher the quality of the study, the more ...	The growth rate over five years will be betwee...	5 in 100 people prior to the intervention to 6...	Vase	The medication increases recovery by 50%	Less than $102	Exactly the same as today with the money in th...	True	1000	10	1.000	59 out of 1000	9.0	20	Crosicol	< INR 500,000	25000.0	41	3	2	1	25000.0	2	2	1	2	2	2	3	1	1	1	2	1	9.0	20	1	1000	10	1.000	3	1	1	0	1	1	0	0	0	1	1	0	0	0	1	0	0	1	0	0
R_401xXJ9E3ZgU9gu	96	R_401xXJ9E3ZgU9gu	Yes, I would like to participate in the study ...	Male	33	Undergraduate Program	Stock analyst	7 (willing to take risk)	8	5	4	5	5	5	6 (strongly agree)	6 (strongly agree)	6 (strongly agree)	0.0	0.0	10	0	30.0	60	0.0	2	1	1	Fingerprint,DNA test	The higher the quality of the study, the more ...	The growth rate over five years will be betwee...	100 in 10,000 people prior to the intervention...	Vase	The medication increases recovery by 100%	More than $102	Less than today with the money in this account	False	500	1	0.001	59 out of 1000	25.0	20	They are equal	INR 500,001 – INR 15,00,000	16000000.0	33	2	2	2	16000000.0	2	1	1	2	2	2	3	2	2	2	1	2	25.0	20	3	500	1	0.001	3	1	0	0	1	1	0	0	1	0	0	1	0	1	1	1	1	0	0	0
R_40N8bUn3C0jwGzf	200	R_40N8bUn3C0jwGzf	Yes, I would like to participate in the study ...	Female	18	Undergraduate Program	Student	0 (unwilling to take risk)	0 (no investment experience)	3	3	4	3	3	3	2	3	40.0	20.0	10	10	10.0	5	5.0	2	1	1	HIV test,Fingerprint,DNA test,Cancer screening...	The higher the quality of the study, the more ...	The growth rate over five years will be betwee...	It is not possible to determine which of the a...	Vase	The medication increases recovery by 2%	More than $102	More than today with the money in this account	True	1000	500	100.000	9 out of 59	11.0	60	They are equal	< INR 500,000	80000.0	18	2	2	1	80000.0	1	1	1	1	2	2	3	2	2	2	3	4	11.0	60	3	1000	500	100.000	1	0	0	0	0	1	0	0	1	0	0	0	1	0	0	1	0	0	0	1
R_40VyTmJ6i96wUP4	235	R_40VyTmJ6i96wUP4	Yes, I would like to participate in the study ...	Female	34	Post-Graduate Program	Not employed	2	0 (no investment experience)	3	2	2	3	3	3	2	3	40.0	25.0	10	10	15.0	0	0.0	1	2	1	DNA test	The higher the quality of the study, the more ...	It is not possible to predict the growth rate ...	It is not possible to determine which of the a...	Vase	The medication increases recovery by 2%	Refuse to answer	Less than today with the money in this account	Do not know	698	352	1.000	59 out of 1000	30.0	20	Crosicol	< INR 500,000	500000.0	34	2	1	1	500000.0	2	2	1	2	2	2	4	1	1	1	3	4	30.0	20	1	698	352	1.000	3	1	1	0	1	1	0	1	0	1	1	0	1	0	1	0	0	0	0	0
R_40ZEg7vX3Y7mMQl	184	R_40ZEg7vX3Y7mMQl	Yes, I would like to participate in the study ...	Female	31	Undergraduate Program	Not employed	4	6	2	2	2	1 (strongly disagree)	2	3	4	1 (strongly disagree)	30.0	30.0	10	0	20.0	10	0.0	1	1	1	HIV test,Fingerprint,DNA test,Cancer screening...	The higher the quality of the study, the more ...	The growth rate will be 0.4% on average each year	70 in 100 people prior to the intervention to ...	Vase	The medication increases recovery by 2%	More than $102	Less than today with the money in this account	True	500	10	10.000	59 out of 1000	50.0	20	They are equal	< INR 500,000	500000.0	31	2	2	1	500000.0	1	1	1	1	2	2	1	1	1	1	3	3	50.0	20	3	500	10	10.000	3	0	0	0	0	1	0	0	0	1	1	0	0	0	1	1	1	1	0	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
R_4yzFrFYMa7w8nQZ	238	R_4yzFrFYMa7w8nQZ	Yes, I would like to participate in the study ...	Female	18	Secondary School (11th to 12th Std.)	Student	0 (unwilling to take risk)	0 (no investment experience)	6 (strongly agree)	5	5	5	4	4	5	6 (strongly agree)	20.0	10.0	15	25	5.0	9	16.0	1	2	1	DNA test	The higher the quality of the study, the more ...	The growth rate over five years will be exactl...	70 in 100 people prior to the intervention to ...	Vase	The medication increases recovery by 100%	More than $102	Less than today with the money in this account	True	1000	100	100.000	59 out of 1000	25.0	20	Hertinol	< INR 500,000	90000.0	18	2	3	1	90000.0	2	2	1	2	2	2	2	1	1	1	1	3	25.0	20	2	1000	100	100.000	3	1	1	0	1	1	0	0	0	1	1	1	0	1	1	0	0	0	0	0
R_4zPiSk23ayGvFD4	364	R_4zPiSk23ayGvFD4	Yes, I would like to participate in the study ...	Male	70	Undergraduate Program	Salaried/Employee/Consultant in a sector other...	4	6	4	4	4	4	5	5	5	4	15.0	20.0	10	0	25.0	30	0.0	2	1	1	DNA test	The higher the quality of the study, the more ...	It is not possible to predict the growth rate ...	It is not possible to determine which of the a...	Vase	The medication increases recovery by 100%	More than $102	Less than today with the money in this account	False	500	10	1.000	59 out of 1000	25.0	20	Crosicol	INR 1500,001 – INR 30,00,000	7000000.0	70	4	2	3	7000000.0	2	2	1	2	2	2	4	2	2	2	1	4	25.0	20	1	500	10	1.000	3	1	1	0	1	1	0	1	1	0	0	1	1	1	1	0	1	1	0	0
R_4zdzjECYeO2A5Ox	255	R_4zdzjECYeO2A5Ox	Yes, I would like to participate in the study ...	Female	19	Secondary School (11th to 12th Std.)	Student	4	2	6 (strongly agree)	4	5	5	5	6 (strongly agree)	3	4	40.0	50.0	0	10	0.0	0	0.0	2	1	1	Fingerprint,DNA test	The lower the quality of the study, the more l...	The growth rate over five years will be exactl...	70 in 100 people prior to the intervention to ...	Vase	None of the above is implied	More than $102	Less than today with the money in this account	Do not know	500	10	0.100	9 out of 59	25.0	20	They are equal	< INR 500,000	800000.0	19	2	3	1	800000.0	2	1	1	2	2	1	2	2	2	2	4	3	25.0	20	3	500	10	0.100	1	1	0	0	1	1	1	0	1	0	0	0	0	1	1	1	1	1	1	1
R_8k6D0jzzHCC5X3Z	369	R_8k6D0jzzHCC5X3Z	Yes, I would like to participate in the study ...	Male	25	Undergraduate Program	Salaried/Employee/Consultant in a sector other...	7 (willing to take risk)	8	6 (strongly agree)	5	5	4	6 (strongly agree)	6 (strongly agree)	6 (strongly agree)	5	5.0	5.0	5	5	30.0	20	30.0	2	1	1	DNA test	The higher the quality of the study, the more ...	It is not possible to predict the growth rate ...	5 in 100 people prior to the intervention to 6...	Vase	The medication increases recovery by 100%	More than $102	Less than today with the money in this account	False	500	500	50.000	59 out of 1000	25.0	10	Hertinol	> INR 75,00,000	8000000.0	25	2	2	6	8000000.0	2	2	1	2	2	2	4	2	2	2	1	1	25.0	10	2	500	500	50.000	3	1	1	0	1	1	0	1	1	0	0	1	0	1	0	0	1	0	0	0
R_8taN3wUPbeZMxzM	2	R_8taN3wUPbeZMxzM	Yes, I would like to participate in the study ...	Male	20	Undergraduate Program	Student	3	0 (no investment experience)	6 (strongly agree)	6 (strongly agree)	4	1 (strongly disagree)	1 (strongly disagree)	2	1 (strongly disagree)	6 (strongly agree)	30.0	40.0	10	10	0.0	5	5.0	1	1	1	Fingerprint	The higher the quality of the study, the more ...	It is not possible to predict the growth rate ...	70 in 100 people prior to the intervention to ...	Vase	The medication increases recovery by 50%	More than $102	More than today with the money in this account	Do not know	1000	10	50.000	9 out of 59	50.0	20	They are equal	< INR 500,000	10000.0	20	2	2	1	10000.0	2	1	2	2	2	2	4	1	1	1	2	3	50.0	20	3	1000	10	50.000	1	1	0	1	1	1	0	1	0	1	1	0	0	0	1	1	0	1	0	1

370 rows × 87 columns

df1 = df1Facet.copy()

df1["Certainty_3"] = df1["certainty1"] + df1["certainty2"] + df1["certainty3"]
df1["RiskComprehension_4"] = df1["riskcalculation1"] + df1["riskcalculation2"] + df1["riskcalculation3"] + df1["riskcalculation4"]
df1["GraphLiteracy_3"] = df1["graph1"] + df1["graph2"] + df1["graph3"]
df1["Numeracy_3"] = df1["numeracy1"] + df1["numeracy2"] + df1["numeracy3"]
df1["Bayesianreasoning_1"] = df1["numeracy4"]
df1["TotalScore_14"] = df1["Certainty_3"] + df1["RiskComprehension_4"] + df1["GraphLiteracy_3"] + df1["Numeracy_3"] + df1["Bayesianreasoning_1"]

df1["Certainty_%"] = df1["Certainty_3"] / 3 * 100
df1["RiskComprehension_%"] = df1["RiskComprehension_4"] / 4 * 100
df1["GraphLiteracy_%"] = df1["GraphLiteracy_3"] / 3 * 100
df1["Numeracy_%"] = df1["Numeracy_3"] / 3 * 100
df1["Bayesianreasoning_%"] = df1["Bayesianreasoning_1"] / 1 * 100

df1["TotalScore_%"] = df1["TotalScore_14"] / 14 * 100


colReq = ["ResponseId", "age", "age_rec", "isced", "income", "wealth", "Certainty_3", "RiskComprehension_4", "GraphLiteracy_3",
          "Numeracy_3", "Bayesianreasoning_1", "Certainty_%", "RiskComprehension_%", "GraphLiteracy_%", "Numeracy_%", "Bayesianreasoning_%",
          "TotalScore_14", "TotalScore_%",]

df2 = df1[colReq].copy()
df2

		ResponseId	age	age_rec	isced	income	wealth	Certainty_3	RiskComprehension_4	GraphLiteracy_3	Numeracy_3	Bayesianreasoning_1	Certainty_%	RiskComprehension_%	GraphLiteracy_%	Numeracy_%	Bayesianreasoning_%	TotalScore_14	TotalScore_%
ResponseId
R_1YkMM2lMB9aEuVL	150	R_1YkMM2lMB9aEuVL	41	3	2	1	25000.0	2	1	1	2	0	66.666667	25.0	33.333333	66.666667	0.0	6	42.857143
R_401xXJ9E3ZgU9gu	96	R_401xXJ9E3ZgU9gu	33	2	2	2	16000000.0	1	1	3	1	1	33.333333	25.0	100.000000	33.333333	100.0	7	50.000000
R_40N8bUn3C0jwGzf	200	R_40N8bUn3C0jwGzf	18	2	2	1	80000.0	0	1	1	1	0	0.000000	25.0	33.333333	33.333333	0.0	3	21.428571
R_40VyTmJ6i96wUP4	235	R_40VyTmJ6i96wUP4	34	2	1	1	500000.0	2	0	1	2	0	66.666667	0.0	33.333333	66.666667	0.0	5	35.714286
R_40ZEg7vX3Y7mMQl	184	R_40ZEg7vX3Y7mMQl	31	2	2	1	500000.0	0	2	2	2	0	0.000000	50.0	66.666667	66.666667	0.0	6	42.857143
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
R_4yzFrFYMa7w8nQZ	238	R_4yzFrFYMa7w8nQZ	18	2	3	1	90000.0	2	0	2	2	1	66.666667	0.0	66.666667	66.666667	100.0	7	50.000000
R_4zPiSk23ayGvFD4	364	R_4zPiSk23ayGvFD4	70	4	2	3	7000000.0	2	2	2	1	1	66.666667	50.0	66.666667	33.333333	100.0	8	57.142857
R_4zdzjECYeO2A5Ox	255	R_4zdzjECYeO2A5Ox	19	2	3	1	800000.0	1	4	3	1	0	33.333333	100.0	100.000000	33.333333	0.0	9	64.285714
R_8k6D0jzzHCC5X3Z	369	R_8k6D0jzzHCC5X3Z	25	2	2	6	8000000.0	2	1	1	1	1	66.666667	25.0	33.333333	33.333333	100.0	6	42.857143
R_8taN3wUPbeZMxzM	2	R_8taN3wUPbeZMxzM	20	2	2	1	10000.0	2	2	2	2	0	66.666667	50.0	66.666667	66.666667	0.0	8	57.142857

370 rows × 18 columns

# The following graph represents frequency of each data point on "TotalScore_14" or the total score out of 19 questions across the sample.

df2['TotalScore_14'].plot(kind = 'hist', xticks = np.arange(0, 15, step=1), xlabel = 'TotalScore_14', title = 'Frequency of Scores' )

<Axes: title={'center': 'Frequency of Scores'}, xlabel='TotalScore_14', ylabel='Frequency'>

# Absolute mean scores for each facet

(df2[['Certainty_3','RiskComprehension_4','GraphLiteracy_3','Numeracy_3','Bayesianreasoning_1','TotalScore_14']].mean(axis = 0)).plot(kind = 'bar', title = 'Absolute mean of scores for above data set')

<Axes: title={'center': 'Absolute mean of scores for above data set'}>

# Normalised mean scores for each facet

df2[['Certainty_%','RiskComprehension_%','GraphLiteracy_%','Numeracy_%','Bayesianreasoning_%','TotalScore_%']].mean(axis = 0).plot(kind = 'bar', title = 'Normalised mean of scores for above data set')

<Axes: title={'center': 'Normalised mean of scores for above data set'}>

# DF3 = Sorted by income

df3 = df2.sort_values(by = 'income')
df3 = df3.reset_index(drop = True)
df3.drop(df3[df3['income'] == 7].index, inplace = True)

df3

	ResponseId	age	age_rec	isced	income	wealth	Certainty_3	RiskComprehension_4	GraphLiteracy_3	Numeracy_3	Bayesianreasoning_1	Certainty_%	RiskComprehension_%	GraphLiteracy_%	Numeracy_%	Bayesianreasoning_%	TotalScore_14	TotalScore_%
0	R_1YkMM2lMB9aEuVL	41	3	2	1	25000.0	2	1	1	2	0	66.666667	25.0	33.333333	66.666667	0.0	6	42.857143
1	R_4TuAjTfgjSCm6mA	29	2	2	1	5000000.0	0	0	3	2	0	0.000000	0.0	100.000000	66.666667	0.0	5	35.714286
2	R_4U1OQvPwcT66nIZ	31	2	2	1	3000000.0	1	3	3	1	0	33.333333	75.0	100.000000	33.333333	0.0	8	57.142857
3	R_4UEtElZbFSSlXJ7	21	2	1	1	100000.0	2	2	3	2	0	66.666667	50.0	100.000000	66.666667	0.0	9	64.285714
4	R_4UhYEPTA8MuBMxr	23	2	2	1	20000000.0	1	1	2	2	0	33.333333	25.0	66.666667	66.666667	0.0	6	42.857143
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
365	R_4gkipJeAyfFWtDH	59	3	1	6	10000000.0	1	2	1	2	0	33.333333	50.0	33.333333	66.666667	0.0	6	42.857143
366	R_48SQkLXmsHzNHIf	25	2	2	6	100.0	2	1	1	2	0	66.666667	25.0	33.333333	66.666667	0.0	6	42.857143
367	R_4aQfSYuVQoXAOIh	30	2	3	6	50000.0	2	1	1	2	0	66.666667	25.0	33.333333	66.666667	0.0	6	42.857143
368	R_4GeNwNOEgwD5rwL	19	2	3	6	400000.0	2	2	1	1	0	66.666667	50.0	33.333333	33.333333	0.0	6	42.857143
369	R_4rGgRh3Y7Grn2f5	20	2	2	6	500000.0	2	0	0	1	0	66.666667	0.0	0.000000	33.333333	0.0	3	21.428571

370 rows × 18 columns

# Absolute mean score for each facet (Sort by Income)

df3[['Certainty_3','RiskComprehension_4','GraphLiteracy_3','Numeracy_3','Bayesianreasoning_1','TotalScore_14']].mean(axis = 0).plot(kind = 'bar', title = 'Absolute mean score for each facet (Sort by Income)')

<Axes: title={'center': 'Absolute mean score for each facet (Sort by Income)'}>

# Normalised mean score for each facet (Sort by Income)

(df3[['Certainty_%','RiskComprehension_%','GraphLiteracy_%','Numeracy_%','Bayesianreasoning_%','TotalScore_%']].mean(axis = 0)).plot(kind = 'bar', title = 'Normalised mean score for each facet (Sort by Income)')

<Axes: title={'center': 'Normalised mean score for each facet (Sort by Income)'}>

# Normalised mean of each facet as a numerical (sorted by income)

df3.groupby('income')[['Certainty_%','RiskComprehension_%','GraphLiteracy_%','Numeracy_%','Bayesianreasoning_%','TotalScore_%']].mean()

	Certainty_%	RiskComprehension_%	GraphLiteracy_%	Numeracy_%	Bayesianreasoning_%	TotalScore_%
income
1	56.281407	46.231156	52.931323	48.073702	29.145729	48.994975
2	47.798742	47.641509	51.257862	47.169811	30.188679	47.102426
3	45.714286	35.714286	48.571429	49.523810	45.714286	44.285714
4	48.888889	46.666667	42.222222	44.444444	40.000000	45.238095
5	44.444444	33.333333	27.777778	44.444444	16.666667	35.714286
6	62.962963	30.555556	22.222222	55.555556	11.111111	39.682540

# Count of responses for each category (sorted by income)

df3.groupby('income')[['ResponseId']].count()

	ResponseId
income
1	199
2	106
3	35
4	15
5	6
6	9

# Absolute mean of Total Facet score line plot (sorted by income)

df3.groupby('income')[['TotalScore_14']].mean().plot( kind = 'line', title = 'Absolute mean of Total Facet score line plot (sorted by income)').legend(loc='center left',bbox_to_anchor=(1.0, 0.5))

<matplotlib.legend.Legend at 0x2aa605a2f90>

# Absolute mean of each Facet score line plot (sorted by income)

df3.groupby('income')[['Certainty_3','RiskComprehension_4','GraphLiteracy_3','Numeracy_3','Bayesianreasoning_1']].mean().plot( kind = 'line', title = 'Absolute Facet scores mean line plot').legend(loc='center left',bbox_to_anchor=(1.0, 0.5))

<matplotlib.legend.Legend at 0x2aa61682790>

# Normalised mean of each Facet score line plot (sorted by income)


df3.groupby('income')[['Certainty_%','RiskComprehension_%','GraphLiteracy_%','Numeracy_%','Bayesianreasoning_%', 'TotalScore_%']].mean().plot( kind = 'line', title = 'Normalised mean of each Facet score line plot (sorted by income)').legend(loc='center left',bbox_to_anchor=(1.0, 0.5))

<matplotlib.legend.Legend at 0x2aa61740c50>

# Normalised mean of each Facet score stacked plot (sorted by income)

df3.groupby('income')[['Certainty_%','RiskComprehension_%','GraphLiteracy_%','Numeracy_%','Bayesianreasoning_%']].mean().plot( kind = 'bar', title = 'Normalised mean of each Facet score stacked plot (sorted by income)', stacked = True).legend(loc='center left',bbox_to_anchor=(1.0, 0.5))

<matplotlib.legend.Legend at 0x2aa60697690>

# Normalised mean of each Facet score hist plot (sorted by income)

df3.groupby('income')[['Certainty_%','RiskComprehension_%','GraphLiteracy_%','Numeracy_%','Bayesianreasoning_%','TotalScore_%']].mean().T.plot(kind = 'bar', title = 'Normalised mean scores for each facet for each wealth response category').legend(loc='center left',bbox_to_anchor=(1.0, 0.5))

<matplotlib.legend.Legend at 0x2aa5ffd18d0>

# Trend line for Absolute Total Facet Score vs income

sns.regplot (data = df3, x = 'income', y = 'TotalScore_14')

<Axes: xlabel='income', ylabel='TotalScore_14'>

# Trend line for Absolute Independent Facet Score vs income reponses

fig, ax = plt.subplots(figsize=(6, 6))

sns.regplot (data = df3, x = 'income', y = 'Certainty_3', fit_reg=True, ci=None, ax=ax, label='Certainty_3')
sns.regplot (data = df3, x = 'income', y = 'RiskComprehension_4', fit_reg=True, ci=None, ax=ax, label='RiskComprehension_4')
sns.regplot (data = df3, x = 'income', y = 'Numeracy_3', fit_reg=True, ci=None, ax=ax, label='Numeracy_3')
sns.regplot (data = df3, x = 'income', y = 'GraphLiteracy_3', fit_reg=True, ci=None, ax=ax, label='GraphLiteracy_3')
sns.regplot (data = df3, x = 'income', y = 'Bayesianreasoning_1',fit_reg=True, ci=None, ax=ax, label='Bayesianreasoning_1' )

ax.set(ylabel='Scores', xlabel='income')
ax.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

# Violine Plot for TotalScore_14 for each entry under income.

sns.violinplot( data = df3, x = 'income', y = 'TotalScore_14')

<Axes: xlabel='income', ylabel='TotalScore_14'>

# Violine Plot for Certainty_3 for each entry under income.

sns.violinplot( data = df3, x = 'income', y = 'Certainty_3')

<Axes: xlabel='income', ylabel='Certainty_3'>

# Violine Plot for RiskComprehension_4 for each entry under income.

sns.violinplot( data = df3, x = 'income', y = 'RiskComprehension_4')

<Axes: xlabel='income', ylabel='RiskComprehension_4'>

# Violine Plot for GraphLiteracy_3 for each entry under income.

sns.violinplot( data = df3, x = 'income', y = 'GraphLiteracy_3')

<Axes: xlabel='income', ylabel='GraphLiteracy_3'>

# Violine Plot for Numeracy_3 for each entry under income.

sns.violinplot( data = df3, x = 'income', y = 'Numeracy_3')

<Axes: xlabel='income', ylabel='Numeracy_3'>

# Violine Plot for TotalScore_14 for each entry under income.

sns.violinplot( data = df3, x = 'income', y = 'Bayesianreasoning_1')

<Axes: xlabel='income', ylabel='Bayesianreasoning_1'>

# Data Frame 2 or df2 is an aggregate data on facet total scores and over all total score, along with wealth and income data vs each response. 
# This data set is also sorted by wealth with all responses 'NA' removed.
# We also assign quartiles ranking each response according to this sort.

df4 = df2.sort_values(by = 'wealth')
df4 = df4.reset_index(drop = True)
df4 = df4.dropna(axis = 0, subset = 'wealth')
df4.insert(loc = len(df4.columns), column = "Quartile Number", value = pd.qcut(df4["wealth"],q = 4, labels = False ) + 1, allow_duplicates = 'False')

df4

	ResponseId	age	age_rec	isced	income	wealth	Certainty_3	RiskComprehension_4	GraphLiteracy_3	Numeracy_3	Bayesianreasoning_1	Certainty_%	RiskComprehension_%	GraphLiteracy_%	Numeracy_%	Bayesianreasoning_%	TotalScore_14	TotalScore_%	Quartile Number
0	R_4DPGUOAP86yO30R	19	2	3	1	0.0	2	2	2	1	0	66.666667	50.0	66.666667	33.333333	0.0	7	50.000000	1
1	R_4lnmGavSf6rw1eE	31	2	1	1	0.0	0	3	1	1	0	0.000000	75.0	33.333333	33.333333	0.0	5	35.714286	1
2	R_4MJAzsNYYA69Y8p	27	2	3	1	0.0	2	1	1	2	1	66.666667	25.0	33.333333	66.666667	100.0	7	50.000000	1
3	R_4fNO2Z5GF8KerZI	24	2	2	2	0.0	3	3	1	2	1	100.000000	75.0	33.333333	66.666667	100.0	10	71.428571	1
4	R_41bfnamM0zpH94i	30	2	1	2	0.0	2	2	1	1	1	66.666667	50.0	33.333333	33.333333	100.0	7	50.000000	1
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
365	R_4M4Yd7LGKym7BaF	32	2	1	3	100000000.0	2	1	3	2	0	66.666667	25.0	100.000000	66.666667	0.0	8	57.142857	4
366	R_4XgkSFVYD7YAAsS	18	2	2	3	100000000.0	1	1	1	2	0	33.333333	25.0	33.333333	66.666667	0.0	5	35.714286	4
367	R_4FZL5zwyR3HMcyl	20	2	3	2	100000000.0	2	2	2	1	0	66.666667	50.0	66.666667	33.333333	0.0	7	50.000000	4
368	R_4f2KZvIaLdZvVqN	27	2	2	5	120000000.0	2	2	2	1	0	66.666667	50.0	66.666667	33.333333	0.0	7	50.000000	4
369	R_4kHu85k7anbQck8	25	2	2	2	150000000.0	0	3	1	1	1	0.000000	75.0	33.333333	33.333333	100.0	6	42.857143	4

370 rows × 19 columns

# Absolute mean score for each facet (Sort by Wealth)

df4[['Certainty_3','RiskComprehension_4','Numeracy_3','GraphLiteracy_3','Bayesianreasoning_1','TotalScore_14']].mean(axis = 0).plot(kind = 'bar', title = 'Absolute mean score for each facet (Sort by Wealth)')

<Axes: title={'center': 'Absolute mean score for each facet (Sort by Wealth)'}>

# Normalised mean score for each facet (Sort by Wealth)

(df4[['Certainty_%','RiskComprehension_%','Numeracy_%','GraphLiteracy_%','Bayesianreasoning_%','TotalScore_%']].mean(axis = 0)).plot(kind = 'bar', title = 'Normalised mean score for each facet (Sort by Wealth)')

<Axes: title={'center': 'Normalised mean score for each facet (Sort by Wealth)'}>

# Line Plot of Absolute Total Facet score vs sorted wealth

df4.plot(use_index = 'True' , kind = 'line', y = ['TotalScore_14'], title = 'Total responses line plot sorted by Wealth',xlabel = 'Unique responses sorted by wealth', ylabel = 'TotalScore_14')

<Axes: title={'center': 'Total responses line plot sorted by Wealth'}, xlabel='Unique responses sorted by wealth', ylabel='TotalScore_14'>

# Absolute Facet scores line plot from low to high wealth 
# I do understand that this is non readable

df4[['Certainty_3','RiskComprehension_4','Numeracy_3','GraphLiteracy_3','Bayesianreasoning_1']].plot( kind = 'line', title = 'Absolute Facet scores line plot', use_index = True).legend(loc='center left',bbox_to_anchor=(1.0, 0.5))

<matplotlib.legend.Legend at 0x2aa60da2210>

# Trend line for Absolute Total Facet Score vs unique wealth reponses

sns.regplot (data = df4, x = df4.index, y = 'TotalScore_14')

<Axes: ylabel='TotalScore_14'>

# Trend line for Absolute Independent Facet Score vs wealth reponses

fig1, ax1 = plt.subplots(figsize=(6, 6))

sns.regplot (data = df4, x = df4.index, y = 'Certainty_3', fit_reg=True, ci=None, ax=ax1, label='Certainty_3')
sns.regplot (data = df4, x = df4.index, y = 'RiskComprehension_4', fit_reg=True, ci=None, ax=ax1, label='RiskComprehension_4')
sns.regplot (data = df4, x = df4.index, y = 'Numeracy_3', fit_reg=True, ci=None, ax=ax1, label='Numeracy_3')
sns.regplot (data = df4, x = df4.index, y = 'GraphLiteracy_3', fit_reg=True, ci=None, ax=ax1, label='GraphLiteracy_3')
sns.regplot (data = df4, x = df4.index, y = 'Bayesianreasoning_1',fit_reg=True, ci=None, ax=ax1, label='Bayesianreasoning_1' )

ax1.set(ylabel='Scores', xlabel='wealth')
ax1.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

# Count of number of responses under each quartile

df4.groupby('Quartile Number')[['Quartile Number']].count()

	Quartile Number
Quartile Number
1	97
2	91
3	96
4	86

df4.loc[df4['Quartile Number']==1]

	ResponseId	age	age_rec	isced	income	wealth	Certainty_3	RiskComprehension_4	GraphLiteracy_3	Numeracy_3	Bayesianreasoning_1	Certainty_%	RiskComprehension_%	GraphLiteracy_%	Numeracy_%	Bayesianreasoning_%	TotalScore_14	TotalScore_%	Quartile Number
0	R_4DPGUOAP86yO30R	19	2	3	1	0.0	2	2	2	1	0	66.666667	50.0	66.666667	33.333333	0.0	7	50.000000	1
1	R_4lnmGavSf6rw1eE	31	2	1	1	0.0	0	3	1	1	0	0.000000	75.0	33.333333	33.333333	0.0	5	35.714286	1
2	R_4MJAzsNYYA69Y8p	27	2	3	1	0.0	2	1	1	2	1	66.666667	25.0	33.333333	66.666667	100.0	7	50.000000	1
3	R_4fNO2Z5GF8KerZI	24	2	2	2	0.0	3	3	1	2	1	100.000000	75.0	33.333333	66.666667	100.0	10	71.428571	1
4	R_41bfnamM0zpH94i	30	2	1	2	0.0	2	2	1	1	1	66.666667	50.0	33.333333	33.333333	100.0	7	50.000000	1
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
92	R_49sPXSg2MnDTzkR	21	2	2	1	100000.0	1	4	3	2	1	33.333333	100.0	100.000000	66.666667	100.0	11	78.571429	1
93	R_49ucUf021OosPGQ	35	2	1	1	100000.0	1	3	3	1	0	33.333333	75.0	100.000000	33.333333	0.0	8	57.142857	1
94	R_4A9Iv5lYYQeK969	36	3	1	2	100000.0	2	1	1	1	0	66.666667	25.0	33.333333	33.333333	0.0	5	35.714286	1
95	R_4CsvgqFBLfzgRSL	43	3	1	1	100000.0	2	2	2	1	0	66.666667	50.0	66.666667	33.333333	0.0	7	50.000000	1
96	R_4CCuDlH0lEcMARo	23	2	2	1	100000.0	0	2	3	1	0	0.000000	50.0	100.000000	33.333333	0.0	6	42.857143	1

97 rows × 19 columns

# This graph is basically a line plot of the TotalScore_14 across first quartile unique responses which are ordered by income

df4.loc[df4['Quartile Number']==1].plot(use_index = 'True' , kind = 'line', y = ['TotalScore_14'], title = 'Total Score line plot sorted by Wealth Q1',xlabel = 'Unique responses', ylabel = 'TotalScore_14')

<Axes: title={'center': 'Total Score line plot sorted by Wealth Q1'}, xlabel='Unique responses', ylabel='TotalScore_14'>

# Trend line for Absolute Total Facet Score vs unique Q14. Income reponses in Quartile 1

sns.regplot (data = df4.loc[df4['Quartile Number']==1], x = df4.loc[df4['Quartile Number']==1].index, y = 'TotalScore_14')

<Axes: ylabel='TotalScore_14'>

# Trend line for Absolute Independent Facet Score vs wealth reponses in Quartile 1

fig1, ax2 = plt.subplots(figsize=(6, 6))

sns.regplot (data = df4.loc[df4['Quartile Number']==1], x = df4.loc[df4['Quartile Number']==1].index, y = 'Certainty_3', fit_reg=True, ci=None, ax=ax2, label='Certainty_3')
sns.regplot (data = df4.loc[df4['Quartile Number']==1], x = df4.loc[df4['Quartile Number']==1].index, y = 'RiskComprehension_4', fit_reg=True, ci=None, ax=ax2, label='RiskComprehension_4')
sns.regplot (data = df4.loc[df4['Quartile Number']==1], x = df4.loc[df4['Quartile Number']==1].index, y = 'Numeracy_3', fit_reg=True, ci=None, ax=ax2, label='Numeracy_3')
sns.regplot (data = df4.loc[df4['Quartile Number']==1], x = df4.loc[df4['Quartile Number']==1].index, y = 'GraphLiteracy_3', fit_reg=True, ci=None, ax=ax2, label='GraphLiteracy_3')
sns.regplot (data = df4.loc[df4['Quartile Number']==1], x = df4.loc[df4['Quartile Number']==1].index, y = 'Bayesianreasoning_1',fit_reg=True, ci=None, ax=ax2, label='Bayesianreasoning_1' )

ax2.set(ylabel='Scores', xlabel='wealth')
ax2.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

df4.loc[df4['Quartile Number']==2]

	ResponseId	age	age_rec	isced	income	wealth	Certainty_3	RiskComprehension_4	GraphLiteracy_3	Numeracy_3	Bayesianreasoning_1	Certainty_%	RiskComprehension_%	GraphLiteracy_%	Numeracy_%	Bayesianreasoning_%	TotalScore_14	TotalScore_%	Quartile Number
97	R_4s4QJ9m9ViKLDPf	19	2	3	1	101000.0	1	0	1	2	0	33.333333	0.0	33.333333	66.666667	0.0	4	28.571429	2
98	R_4qh6umVAgcFqF0h	22	2	2	1	112378.0	2	3	3	1	0	66.666667	75.0	100.000000	33.333333	0.0	9	64.285714	2
99	R_4kN4Jn7Y8hHdzlY	23	2	2	1	120000.0	2	0	2	1	0	66.666667	0.0	66.666667	33.333333	0.0	5	35.714286	2
100	R_437fyX4DcVGhqJe	40	3	1	2	150000.0	2	2	1	1	0	66.666667	50.0	33.333333	33.333333	0.0	6	42.857143	2
101	R_4PFudUMcC09ZJk8	36	3	2	2	150000.0	2	3	3	1	0	66.666667	75.0	100.000000	33.333333	0.0	9	64.285714	2
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
183	R_4rGgRh3Y7Grn2f5	20	2	2	6	500000.0	2	0	0	1	0	66.666667	0.0	0.000000	33.333333	0.0	3	21.428571	2
184	R_40VyTmJ6i96wUP4	34	2	1	1	500000.0	2	0	1	2	0	66.666667	0.0	33.333333	66.666667	0.0	5	35.714286	2
185	R_4eUXMOw0QY7ovqu	24	2	2	3	500000.0	0	2	1	2	1	0.000000	50.0	33.333333	66.666667	100.0	6	42.857143	2
186	R_4C2WItrJsoTF1cE	19	2	2	1	500000.0	2	4	1	2	1	66.666667	100.0	33.333333	66.666667	100.0	10	71.428571	2
187	R_4wdhk4DDjmecH1A	29	2	2	2	500000.0	0	1	2	1	1	0.000000	25.0	66.666667	33.333333	100.0	5	35.714286	2

91 rows × 19 columns

# This graph is basically a line plot of the TotalScore_14 across second quartile unique responses which are ordered by income

df4.loc[df4['Quartile Number']==2].plot(use_index = 'True' , kind = 'line', y = ['TotalScore_14'], title = 'Total Score line plot sorted by Wealth Q1',xlabel = 'Unique responses', ylabel = 'TotalScore_14')

<Axes: title={'center': 'Total Score line plot sorted by Wealth Q1'}, xlabel='Unique responses', ylabel='TotalScore_14'>

# Trend line for Absolute Total Facet Score vs unique wealth reponses in Quartile 2

sns.regplot (data = df4.loc[df4['Quartile Number']==2], x = df4.loc[df4['Quartile Number']==2].index, y = 'TotalScore_14')

<Axes: ylabel='TotalScore_14'>

# Trend line for Absolute Independent Facet Score vs wealth reponses in Quartile 2

fig1, ax3 = plt.subplots(figsize=(6, 6))

sns.regplot (data = df4.loc[df4['Quartile Number']==2], x = df4.loc[df4['Quartile Number']==2].index, y = 'Certainty_3', fit_reg=True, ci=None, ax=ax3, label='Certainty_3')
sns.regplot (data = df4.loc[df4['Quartile Number']==2], x = df4.loc[df4['Quartile Number']==2].index, y = 'RiskComprehension_4', fit_reg=True, ci=None, ax=ax3, label='RiskComprehension_4')
sns.regplot (data = df4.loc[df4['Quartile Number']==2], x = df4.loc[df4['Quartile Number']==2].index, y = 'Numeracy_3', fit_reg=True, ci=None, ax=ax3, label='Numeracy_3')
sns.regplot (data = df4.loc[df4['Quartile Number']==2], x = df4.loc[df4['Quartile Number']==2].index, y = 'GraphLiteracy_3', fit_reg=True, ci=None, ax=ax3, label='GraphLiteracy_3')
sns.regplot (data = df4.loc[df4['Quartile Number']==2], x = df4.loc[df4['Quartile Number']==2].index, y = 'Bayesianreasoning_1',fit_reg=True, ci=None, ax=ax3, label='Bayesianreasoning_1' )

ax3.set(ylabel='Scores', xlabel='wealth')
ax3.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

df4.loc[df4['Quartile Number']==3]

	ResponseId	age	age_rec	isced	income	wealth	Certainty_3	RiskComprehension_4	GraphLiteracy_3	Numeracy_3	Bayesianreasoning_1	Certainty_%	RiskComprehension_%	GraphLiteracy_%	Numeracy_%	Bayesianreasoning_%	TotalScore_14	TotalScore_%	Quartile Number
188	R_4MbebluzsBWokIc	57	3	2	2	550000.0	1	3	2	1	0	33.333333	75.0	66.666667	33.333333	0.0	7	50.000000	3
189	R_4TuHdn6ifelyTqF	42	3	2	2	600000.0	2	4	2	1	1	66.666667	100.0	66.666667	33.333333	100.0	10	71.428571	3
190	R_4Issx8w7LuZg7Yt	28	2	2	2	600000.0	2	2	1	2	0	66.666667	50.0	33.333333	66.666667	0.0	7	50.000000	3
191	R_4jTmQ2fyl4GIZKf	24	2	2	1	600000.0	2	2	2	1	0	66.666667	50.0	66.666667	33.333333	0.0	7	50.000000	3
192	R_4EpvHZs1AkqudDX	24	2	2	1	600000.0	2	3	3	1	1	66.666667	75.0	100.000000	33.333333	100.0	10	71.428571	3
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
279	R_4IXCBojUC4PBdve	51	3	2	2	3000000.0	2	2	3	2	1	66.666667	50.0	100.000000	66.666667	100.0	10	71.428571	3
280	R_4EcDVJCgRS2diGg	32	2	1	5	3000000.0	0	2	2	1	1	0.000000	50.0	66.666667	33.333333	100.0	6	42.857143	3
281	R_41hdf8aceebqOfl	25	2	2	2	3000000.0	2	3	1	2	0	66.666667	75.0	33.333333	66.666667	0.0	8	57.142857	3
282	R_4EneslabJXcCOT4	18	2	3	1	3000000.0	2	3	2	1	0	66.666667	75.0	66.666667	33.333333	0.0	8	57.142857	3
283	R_4pPBZhFSJIbPPha	45	3	1	3	3000000.0	0	3	2	1	1	0.000000	75.0	66.666667	33.333333	100.0	7	50.000000	3

96 rows × 19 columns

# This graph is basically a line plot of the TotalScore_14 across third quartile unique responses which are ordered by wealth

df4.loc[df4['Quartile Number']==3].plot(use_index = 'True' , kind = 'line', y = ['TotalScore_14'], title = 'Total Score line plot sorted by Wealth Q3',xlabel = 'Unique responses', ylabel = 'TotalScore_14')

<Axes: title={'center': 'Total Score line plot sorted by Wealth Q3'}, xlabel='Unique responses', ylabel='TotalScore_14'>

# Trend line for Absolute Total Facet Score vs unique wealth reponses in Quartile 3

sns.regplot (data = df4.loc[df4['Quartile Number']==3], x = df4.loc[df4['Quartile Number']==3].index, y = 'TotalScore_14')

<Axes: ylabel='TotalScore_14'>

# Trend line for Absolute Independent Facet Score vs wealth reponses in Quartile 3

fig1, ax4 = plt.subplots(figsize=(6, 6))

sns.regplot (data = df4.loc[df4['Quartile Number']==3], x = df4.loc[df4['Quartile Number']==3].index, y = 'Certainty_3', fit_reg=True, ci=None, ax=ax4, label='Certainty_3')
sns.regplot (data = df4.loc[df4['Quartile Number']==3], x = df4.loc[df4['Quartile Number']==3].index, y = 'RiskComprehension_4', fit_reg=True, ci=None, ax=ax4, label='RiskComprehension_4')
sns.regplot (data = df4.loc[df4['Quartile Number']==3], x = df4.loc[df4['Quartile Number']==3].index, y = 'Numeracy_3', fit_reg=True, ci=None, ax=ax4, label='Numeracy_3')
sns.regplot (data = df4.loc[df4['Quartile Number']==3], x = df4.loc[df4['Quartile Number']==3].index, y = 'GraphLiteracy_3', fit_reg=True, ci=None, ax=ax4, label='GraphLiteracy_3')
sns.regplot (data = df4.loc[df4['Quartile Number']==3], x = df4.loc[df4['Quartile Number']==3].index, y = 'Bayesianreasoning_1',fit_reg=True, ci=None, ax=ax4, label='Bayesianreasoning_1' )

ax4.set(ylabel='Scores', xlabel='wealth')
ax4.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

df4.loc[df4['Quartile Number']==4]

	ResponseId	age	age_rec	isced	income	wealth	Certainty_3	RiskComprehension_4	GraphLiteracy_3	Numeracy_3	Bayesianreasoning_1	Certainty_%	RiskComprehension_%	GraphLiteracy_%	Numeracy_%	Bayesianreasoning_%	TotalScore_14	TotalScore_%	Quartile Number
284	R_4FVQvDpbnyvqOQY	28	2	1	2	3200000.0	0	1	0	2	0	0.000000	25.0	0.000000	66.666667	0.0	3	21.428571	4
285	R_4dyDxqwOf6c5bUt	46	3	1	2	4000000.0	2	3	2	1	0	66.666667	75.0	66.666667	33.333333	0.0	8	57.142857	4
286	R_4w1ruZDnw0ZQRxf	22	2	3	4	4000000.0	0	1	1	2	1	0.000000	25.0	33.333333	66.666667	100.0	5	35.714286	4
287	R_4Rft74IKSrfwItv	24	2	1	1	4000000.0	1	4	2	2	1	33.333333	100.0	66.666667	66.666667	100.0	10	71.428571	4
288	R_4NERL55GtGXfKj7	40	3	2	1	4000000.0	2	1	2	2	1	66.666667	25.0	66.666667	66.666667	100.0	8	57.142857	4
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
365	R_4M4Yd7LGKym7BaF	32	2	1	3	100000000.0	2	1	3	2	0	66.666667	25.0	100.000000	66.666667	0.0	8	57.142857	4
366	R_4XgkSFVYD7YAAsS	18	2	2	3	100000000.0	1	1	1	2	0	33.333333	25.0	33.333333	66.666667	0.0	5	35.714286	4
367	R_4FZL5zwyR3HMcyl	20	2	3	2	100000000.0	2	2	2	1	0	66.666667	50.0	66.666667	33.333333	0.0	7	50.000000	4
368	R_4f2KZvIaLdZvVqN	27	2	2	5	120000000.0	2	2	2	1	0	66.666667	50.0	66.666667	33.333333	0.0	7	50.000000	4
369	R_4kHu85k7anbQck8	25	2	2	2	150000000.0	0	3	1	1	1	0.000000	75.0	33.333333	33.333333	100.0	6	42.857143	4

86 rows × 19 columns

# This graph is basically a line plot of the TotalScore_14 across fourth quartile unique responses which are ordered by Wealth

df4.loc[df4['Quartile Number']==4].plot(use_index = 'True' , kind = 'line', y = ['TotalScore_14'], title = 'Total Score line plot sorted by Wealth Q4',xlabel = 'Unique responses', ylabel = 'TotalScore_14')

<Axes: title={'center': 'Total Score line plot sorted by Wealth Q4'}, xlabel='Unique responses', ylabel='TotalScore_14'>

# Trend line for Absolute Total Facet Score vs unique wealth reponses in Quartile 4

sns.regplot (data = df4.loc[df4['Quartile Number']==4], x = df4.loc[df4['Quartile Number']==4].index, y = 'TotalScore_14')

<Axes: ylabel='TotalScore_14'>

# Trend line for Absolute Independent Facet Score vs wealth reponses in Quartile 4

fig1, ax5 = plt.subplots(figsize=(6, 6))

sns.regplot (data = df4.loc[df4['Quartile Number']==4], x = df4.loc[df4['Quartile Number']==4].index, y = 'Certainty_3', fit_reg=True, ci=None, ax=ax5, label='Certainty_3')
sns.regplot (data = df4.loc[df4['Quartile Number']==4], x = df4.loc[df4['Quartile Number']==4].index, y = 'RiskComprehension_4', fit_reg=True, ci=None, ax=ax5, label='RiskComprehension_4')
sns.regplot (data = df4.loc[df4['Quartile Number']==4], x = df4.loc[df4['Quartile Number']==4].index, y = 'Numeracy_3', fit_reg=True, ci=None, ax=ax5, label='Numeracy_3')
sns.regplot (data = df4.loc[df4['Quartile Number']==4], x = df4.loc[df4['Quartile Number']==4].index, y = 'GraphLiteracy_3', fit_reg=True, ci=None, ax=ax5, label='GraphLiteracy_3')
sns.regplot (data = df4.loc[df4['Quartile Number']==4], x = df4.loc[df4['Quartile Number']==4].index, y = 'Bayesianreasoning_1',fit_reg=True, ci=None, ax=ax5, label='Bayesianreasoning_1' )

ax5.set(ylabel='Scores', xlabel='wealth')
ax5.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

# Absolute Total Facet scores mean line plot from low to high Wealth

df4.groupby('Quartile Number')[['TotalScore_14']].mean().plot( kind = 'line', title = 'Absolute Total Facet scores mean line plot low to high wealth', xticks = np.arange(1,5,step = 1)).legend(loc='center left',bbox_to_anchor=(1.0, 0.5))

<matplotlib.legend.Legend at 0x2aa6018aa90>

# Absolute Facet scores mean line plot from low to high wealth

df4.groupby('Quartile Number')[['Certainty_3','RiskComprehension_4','Numeracy_3','GraphLiteracy_3','Bayesianreasoning_1']].mean().plot( kind = 'line', title = 'Absolute Facet scores mean line plot low to high wealth', xticks = np.arange(1,5,step = 1)).legend(loc='center left',bbox_to_anchor=(1.0, 0.5))

<matplotlib.legend.Legend at 0x2aa60ce4d50>

# Normalised Facet scores mean line plot from low to high wealth


df4.groupby('Quartile Number')[['Certainty_%','RiskComprehension_%','Numeracy_%','GraphLiteracy_%','Bayesianreasoning_%','TotalScore_%']].mean().plot( kind = 'line', title = 'Normalised Facet scores mean line plot from low to high wealth',  xticks = np.arange(1,5,step = 1)).legend(loc='center left',bbox_to_anchor=(1.0, 0.5))

<matplotlib.legend.Legend at 0x2aa60343190>

# Normalised mean scores for each facet stacked

df4.groupby('Quartile Number')[['Certainty_%','RiskComprehension_%','Numeracy_%','GraphLiteracy_%','Bayesianreasoning_%']].mean().plot( kind = 'bar', title = 'Normalised mean scores for each facet stacked', stacked = True).legend(loc='center left',bbox_to_anchor=(1.0, 0.5))

<matplotlib.legend.Legend at 0x2aa60307310>

# Normalised mean scores for each facet for each wealth response Quartile

df4.groupby('Quartile Number')[['Certainty_%','RiskComprehension_%','Numeracy_%','GraphLiteracy_%','Bayesianreasoning_%','TotalScore_%']].mean().T.plot(kind = 'bar', title = 'Normalised mean scores for each facet for each wealth response category').legend(loc='center left',bbox_to_anchor=(1.0, 0.5))

<matplotlib.legend.Legend at 0x2aa602fffd0>

# Trend line for Absolute Total Facet Score vs Quartiles

sns.regplot (data = df4, x = 'Quartile Number', y = 'TotalScore_14')

<Axes: xlabel='Quartile Number', ylabel='TotalScore_14'>

# Trend line for Absolute Independent Facet Score vs Quartiles

fig2, ax2 = plt.subplots(figsize=(6, 6))

sns.regplot (data = df4, x = 'Quartile Number', y = 'Certainty_3', fit_reg=True, ci=None, ax=ax2, label='Certainty_3')
sns.regplot (data = df4, x = 'Quartile Number', y = 'RiskComprehension_4', fit_reg=True, ci=None, ax=ax2, label='RiskComprehension_4')
sns.regplot (data = df4, x = 'Quartile Number', y = 'Numeracy_3', fit_reg=True, ci=None, ax=ax2, label='Numeracy_3')
sns.regplot (data = df4, x = 'Quartile Number', y = 'GraphLiteracy_3', fit_reg=True, ci=None, ax=ax2, label='GraphLiteracy_3')
sns.regplot (data = df4, x = 'Quartile Number', y = 'Bayesianreasoning_1',fit_reg=True, ci=None, ax=ax2, label='Bayesianreasoning_1' )

ax2.set(ylabel='Scores', xlabel='wealth')
ax2.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

# Violine Plot for TotalScore_14 for each entry under Wealth Quartiles.

sns.violinplot( data = df4, x = 'Quartile Number', y = 'TotalScore_14')

<Axes: xlabel='Quartile Number', ylabel='TotalScore_14'>

# Violine Plot for Certainty_3 for each entry under Wealth Quartiles.

sns.violinplot( data = df4, x = 'Quartile Number', y = 'Certainty_3')

<Axes: xlabel='Quartile Number', ylabel='Certainty_3'>

# Violine Plot for RiskComprehension_4 for each entry under Wealth Quartiles.

sns.violinplot( data = df4, x = 'Quartile Number', y = 'RiskComprehension_4')

<Axes: xlabel='Quartile Number', ylabel='RiskComprehension_4'>

# Violine Plot for GraphLiteracy_3 for each entry under Wealth Quartiles.

sns.violinplot( data = df4, x = 'Quartile Number', y = 'GraphLiteracy_3')

<Axes: xlabel='Quartile Number', ylabel='GraphLiteracy_3'>

# Violine Plot for Numeracy_3 for each entry under Wealth Quartiles.

sns.violinplot( data = df4, x = 'Quartile Number', y = 'Numeracy_3')

<Axes: xlabel='Quartile Number', ylabel='Numeracy_3'>

# Violine Plot for Bayesianreasoning_1 for each entry under Wealth Quartiles.

sns.violinplot( data = df4, x = 'Quartile Number', y = 'Bayesianreasoning_1')

<Axes: xlabel='Quartile Number', ylabel='Bayesianreasoning_1'>

# Data Frame 5 or df5 is an aggregate data on facet total scores and over all total score, along with  education data vs each response. 
# This data set is now sorted by education.

df5 = df2.sort_values(by = 'isced')
df5 = df5.reset_index(drop = True)

df5

	ResponseId	age	age_rec	isced	income	wealth	Certainty_3	RiskComprehension_4	GraphLiteracy_3	Numeracy_3	Bayesianreasoning_1	Certainty_%	RiskComprehension_%	GraphLiteracy_%	Numeracy_%	Bayesianreasoning_%	TotalScore_14	TotalScore_%
0	R_4XhfVqNhg49zZXe	37	3	1	2	6000000.0	2	0	1	2	0	66.666667	0.0	33.333333	66.666667	0.0	5	35.714286
1	R_4FjN0R9SXA0s0G5	53	3	1	2	5000000.0	2	2	1	1	1	66.666667	50.0	33.333333	33.333333	100.0	7	50.000000
2	R_4Ph1UNtYnaoPSX7	25	2	1	3	400000.0	0	2	0	1	0	0.000000	50.0	0.000000	33.333333	0.0	3	21.428571
3	R_4FVQvDpbnyvqOQY	28	2	1	2	3200000.0	0	1	0	2	0	0.000000	25.0	0.000000	66.666667	0.0	3	21.428571
4	R_4PjnTbc2vIjogot	31	2	1	1	200000.0	2	1	0	1	0	66.666667	25.0	0.000000	33.333333	0.0	4	28.571429
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
365	R_4fkCPjfYl6UggHL	18	2	3	1	1000000.0	2	3	1	2	1	66.666667	75.0	33.333333	66.666667	100.0	9	64.285714
366	R_4K6WL8dnMy7ODbs	18	2	3	6	100000.0	2	0	1	2	0	66.666667	0.0	33.333333	66.666667	0.0	5	35.714286
367	R_4ANgmQQx8TBl3kc	21	2	3	1	16000000.0	0	4	2	1	1	0.000000	100.0	66.666667	33.333333	100.0	8	57.142857
368	R_4MJAzsNYYA69Y8p	27	2	3	1	0.0	2	1	1	2	1	66.666667	25.0	33.333333	66.666667	100.0	7	50.000000
369	R_4EneslabJXcCOT4	18	2	3	1	3000000.0	2	3	2	1	0	66.666667	75.0	66.666667	33.333333	0.0	8	57.142857

370 rows × 18 columns

# since we didnt drop any values, the dataset is essentially the same as df2, but sorted

# Mean of each facet as a % for each category of isced

df5.groupby('isced')[['Certainty_%','RiskComprehension_%','Numeracy_%','GraphLiteracy_%','Bayesianreasoning_%']].mean()

	Certainty_%	RiskComprehension_%	Numeracy_%	GraphLiteracy_%	Bayesianreasoning_%
isced
1	51.851852	44.444444	48.677249	49.470899	28.571429
2	51.741294	46.268657	47.097844	51.741294	31.840796
3	58.139535	41.279070	49.612403	47.286822	32.558140

# Count of responses for each category isced

df5.groupby('isced')[['ResponseId']].count()

	ResponseId
isced
1	126
2	201
3	43

# Absolute Total Facet scores mean line plot from low to high education

df5.groupby('isced')[['TotalScore_14']].mean().plot( kind = 'line', title = 'Absolute Total Facet scores mean line plot', xticks = np.arange(1,4, step = 1)).legend(loc='center left',bbox_to_anchor=(1.0, 0.5))

<matplotlib.legend.Legend at 0x2aa60e82c90>

# Absolute Facet scores mean line plot from low to high edu

df5.groupby('isced')[['Certainty_3','RiskComprehension_4','Numeracy_3','GraphLiteracy_3','Bayesianreasoning_1']].mean().plot( kind = 'line', title = 'Absolute Facet scores mean line plot', xticks = np.arange(1,4, step = 1)).legend(loc='center left',bbox_to_anchor=(1.0, 0.5))

<matplotlib.legend.Legend at 0x2aa67800950>

# Normalised Facet scores mean line plot from low to high edu


df5.groupby('isced')[['Certainty_%','RiskComprehension_%','Numeracy_%','GraphLiteracy_%','Bayesianreasoning_%','TotalScore_%']].mean().plot( kind = 'line', title = 'Normalised Facet scores mean line plot', xticks = np.arange(1,4, step = 1)).legend(loc='center left',bbox_to_anchor=(1.0, 0.5))

<matplotlib.legend.Legend at 0x2aa67803110>

# Normalised mean scores for each facet stacked

df5.groupby('isced')[['Certainty_%','RiskComprehension_%','Numeracy_%','GraphLiteracy_%','Bayesianreasoning_%']].mean().plot( kind = 'bar', title = 'Normalised mean scores for each facet stacked', stacked = True).legend(loc='center left',bbox_to_anchor=(1.0, 0.5))

<matplotlib.legend.Legend at 0x2aa6589a790>

# Normalised mean scores for each facet for each edu response category

df5.groupby('isced')[['Certainty_%','RiskComprehension_%','Numeracy_%','GraphLiteracy_%','Bayesianreasoning_%','TotalScore_%']].mean().T.plot(kind = 'bar', title = 'Normalised mean scores for each facet for each edu response category').legend(loc='center left',bbox_to_anchor=(1.0, 0.5))

<matplotlib.legend.Legend at 0x2aa658d0b90>

# Trend line for Absolute Total Facet Score vs edu (isced) reponses

sns.regplot (data = df5, x = 'isced', y = 'TotalScore_14')

<Axes: xlabel='isced', ylabel='TotalScore_14'>

# Trend line for Absolute Independent Facet Score vs edu (isced) reponses

fig, ax6 = plt.subplots(figsize=(6, 6))

sns.regplot (data = df5, x = 'isced', y = 'Certainty_3', fit_reg=True, ci=None, ax=ax6, label='Certainty_3')
sns.regplot (data = df5, x = 'isced', y = 'RiskComprehension_4', fit_reg=True, ci=None, ax=ax6, label='RiskComprehension_4')
sns.regplot (data = df5, x = 'isced', y = 'Numeracy_3', fit_reg=True, ci=None, ax=ax6, label='Numeracy_3')
sns.regplot (data = df5, x = 'isced', y = 'GraphLiteracy_3', fit_reg=True, ci=None, ax=ax6, label='GraphLiteracy_3')
sns.regplot (data = df5, x = 'isced', y = 'Bayesianreasoning_1',fit_reg=True, ci=None, ax=ax6, label='Bayesianreasoning_1' )

ax6.set(ylabel='Scores', xlabel='isced')
ax6.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

# Violine Plot for TotalScore_14 for each entry under edu.

sns.violinplot( data = df5, x = 'isced', y = 'TotalScore_14')

<Axes: xlabel='isced', ylabel='TotalScore_14'>

# Violine Plot for Certainty_3 for each entry under edu.

sns.violinplot( data = df5, x = 'isced', y = 'Certainty_3')

<Axes: xlabel='isced', ylabel='Certainty_3'>

# Violine Plot for RiskComprehension_4 for each entry under edu.

sns.violinplot( data = df5, x = 'isced', y = 'RiskComprehension_4')

<Axes: xlabel='isced', ylabel='RiskComprehension_4'>

# Violine Plot for GraphLiteracy_3 for each entry under edu.

sns.violinplot( data = df5, x = 'isced', y = 'GraphLiteracy_3')

<Axes: xlabel='isced', ylabel='GraphLiteracy_3'>

# Violine Plot for Numeracy_3 for each entry under edu.

sns.violinplot( data = df5, x = 'isced', y = 'Numeracy_3')

<Axes: xlabel='isced', ylabel='Numeracy_3'>

# Violine Plot for TotalScore_14 for each entry under edu.

sns.violinplot( data = df5, x = 'isced', y = 'Bayesianreasoning_1')

<Axes: xlabel='isced', ylabel='Bayesianreasoning_1'>

# Descriptive stats for the data set, isced = 1
# NA values of wealth are removed

df6 = df5.dropna(axis = 0, subset = 'wealth')
# df6.drop(df6[df6['wealth'] == 3500000].index, inplace = True)
df6.loc[df6['isced']==1][['wealth']].describe()

	wealth
count	1.260000e+02
mean	4.726376e+06
std	1.304076e+07
min	0.000000e+00
25%	2.000000e+05
50%	1.000000e+06
75%	4.000000e+06
max	1.000000e+08

# Descriptive stats for the data set, isced = 2

df6.loc[df6['isced']==2][['wealth']].describe()

	wealth
count	2.010000e+02
mean	5.055470e+06
std	1.647852e+07
min	0.000000e+00
25%	1.000000e+05
50%	5.000000e+05
75%	2.500000e+06
max	1.500000e+08

# Descriptive stats for the data set, isced = 3

df6.loc[df6['isced']==3][['wealth']].describe()

	wealth
count	4.300000e+01
mean	4.133750e+06
std	1.532675e+07
min	0.000000e+00
25%	7.000000e+04
50%	4.000000e+05
75%	1.750000e+06
max	1.000000e+08

# Trend line for Absolute Tota Facet Score vs wealth reponses sorted by isced and ORDERED by wealth WITH scatter

fig, ax7 = plt.subplots(figsize=(6, 6))

sns.regplot (data = df6.loc[df6['isced']==1].reset_index(drop = True), x = df6.loc[df6['isced']==1].reset_index(drop = True)['wealth'], y = 'TotalScore_14', fit_reg=True, ci=None, ax=ax7, label='ISCED = 1')
sns.regplot (data = df6.loc[df6['isced']==2].reset_index(drop = True), x = df6.loc[df6['isced']==2].reset_index(drop = True)['wealth'], y = 'TotalScore_14', fit_reg=True, ci=None, ax=ax7, label='ISCED = 2')
sns.regplot (data = df6.loc[df6['isced']==3].reset_index(drop = True), x = df6.loc[df6['isced']==3].reset_index(drop = True)['wealth'], y = 'TotalScore_14', fit_reg=True, ci=None, ax=ax7, label='ISCED = 3')

ax7.set(ylabel='Total Scores_19', xlabel='Wealth')
ax7.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

# Trend line for Absolute Tota Facet Score vs wealth reponses sorted by isced and ORDERED by wealth WITHOUT scatter

fig, ax8 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df6.loc[df6['isced']==1].reset_index(drop = True), x = df6.loc[df6['isced']==1].reset_index(drop = True)['wealth'], y = 'TotalScore_14', fit_reg=True, ci=None, ax=ax8, label='ISCED = 1')
sns.regplot (scatter = False, data = df6.loc[df6['isced']==2].reset_index(drop = True), x = df6.loc[df6['isced']==2].reset_index(drop = True)['wealth'], y = 'TotalScore_14', fit_reg=True, ci=None, ax=ax8, label='ISCED = 2')
sns.regplot (scatter = False, data = df6.loc[df6['isced']==3].reset_index(drop = True), x = df6.loc[df6['isced']==3].reset_index(drop = True)['wealth'], y = 'TotalScore_14', fit_reg=True, ci=None, ax=ax8, label='ISCED = 3')

ax8.set(ylabel='Total Scores_19', xlabel='Wealth')
ax8.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

# Trend line for Absolute Certainty Score vs wealth reponses sorted by isced and ORDERED by wealth WITHOUT scatter

fig, ax9 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df6.loc[df6['isced']==1].reset_index(drop = True), x = df6.loc[df6['isced']==1].reset_index(drop = True)['wealth'], y = 'Certainty_3', fit_reg=True, ci=None, ax=ax9, label='ISCED = 1')
sns.regplot (scatter = False, data = df6.loc[df6['isced']==2].reset_index(drop = True), x = df6.loc[df6['isced']==2].reset_index(drop = True)['wealth'], y = 'Certainty_3', fit_reg=True, ci=None, ax=ax9, label='ISCED = 2')
sns.regplot (scatter = False, data = df6.loc[df6['isced']==3].reset_index(drop = True), x = df6.loc[df6['isced']==3].reset_index(drop = True)['wealth'], y = 'Certainty_3', fit_reg=True, ci=None, ax=ax9, label='ISCED = 3')

ax9.set(ylabel='Certainty_3', xlabel='Wealth')
ax9.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

# Trend line for Absolute Uncertainty Score vs wealth reponses sorted by isced and ORDERED by wealth WITHOUT scatter

fig, ax10 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df6.loc[df6['isced']==1].reset_index(drop = True), x = df6.loc[df6['isced']==1].reset_index(drop = True)['wealth'], y = 'RiskComprehension_4', fit_reg=True, ci=None, ax=ax10, label='ISCED = 1')
sns.regplot (scatter = False, data = df6.loc[df6['isced']==2].reset_index(drop = True), x = df6.loc[df6['isced']==2].reset_index(drop = True)['wealth'], y = 'RiskComprehension_4', fit_reg=True, ci=None, ax=ax10, label='ISCED = 2')
sns.regplot (scatter = False, data = df6.loc[df6['isced']==3].reset_index(drop = True), x = df6.loc[df6['isced']==3].reset_index(drop = True)['wealth'], y = 'RiskComprehension_4', fit_reg=True, ci=None, ax=ax10, label='ISCED = 3')

ax10.set(ylabel='RiskComprehension_4', xlabel='Wealth')
ax10.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

# Trend line for Absolute Number Comprehension Score vs wealth reponses sorted by isced and ORDERED by wealth WITHOUT scatter

fig, ax11 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df6.loc[df6['isced']==1].reset_index(drop = True), x = df6.loc[df6['isced']==1].reset_index(drop = True)['wealth'], y = 'Numeracy_3', fit_reg=True, ci=None, ax=ax11, label='ISCED = 1')
sns.regplot (scatter = False, data = df6.loc[df6['isced']==2].reset_index(drop = True), x = df6.loc[df6['isced']==2].reset_index(drop = True)['wealth'], y = 'Numeracy_3', fit_reg=True, ci=None, ax=ax11, label='ISCED = 2')
sns.regplot (scatter = False, data = df6.loc[df6['isced']==3].reset_index(drop = True), x = df6.loc[df6['isced']==3].reset_index(drop = True)['wealth'], y = 'Numeracy_3', fit_reg=True, ci=None, ax=ax11, label='ISCED = 3')

ax11.set(ylabel='Numeracy_3', xlabel='Wealth')
ax11.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

# Trend line for Absolute Graph Comprehension Score vs wealth reponses sorted by isced and ORDERED by wealth WITHOUT scatter

fig, ax11 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df6.loc[df6['isced']==1].reset_index(drop = True), x = df6.loc[df6['isced']==1].reset_index(drop = True)['wealth'], y = 'GraphLiteracy_3', fit_reg=True, ci=None, ax=ax11, label='ISCED = 1')
sns.regplot (scatter = False, data = df6.loc[df6['isced']==2].reset_index(drop = True), x = df6.loc[df6['isced']==2].reset_index(drop = True)['wealth'], y = 'GraphLiteracy_3', fit_reg=True, ci=None, ax=ax11, label='ISCED = 2')
sns.regplot (scatter = False, data = df6.loc[df6['isced']==3].reset_index(drop = True), x = df6.loc[df6['isced']==3].reset_index(drop = True)['wealth'], y = 'GraphLiteracy_3', fit_reg=True, ci=None, ax=ax11, label='ISCED = 3')

ax11.set(ylabel='GraphLiteracy_3', xlabel='Wealth')
ax11.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

# Trend line for Absolute Bayesian Reasoning Score vs wealth reponses sorted by isced and ORDERED by wealth WITHOUT scatter

fig, ax12 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df6.loc[df6['isced']==1].reset_index(drop = True), x = df6.loc[df6['isced']==1].reset_index(drop = True)['wealth'], y = 'Bayesianreasoning_1', fit_reg=True, ci=None, ax=ax12, label='ISCED = 1')
sns.regplot (scatter = False, data = df6.loc[df6['isced']==2].reset_index(drop = True), x = df6.loc[df6['isced']==2].reset_index(drop = True)['wealth'], y = 'Bayesianreasoning_1', fit_reg=True, ci=None, ax=ax12, label='ISCED = 2')
sns.regplot (scatter = False, data = df6.loc[df6['isced']==3].reset_index(drop = True), x = df6.loc[df6['isced']==3].reset_index(drop = True)['wealth'], y = 'Bayesianreasoning_1', fit_reg=True, ci=None, ax=ax12, label='ISCED = 3')

ax12.set(ylabel='Bayesianreasoning_1', xlabel='Wealth')
ax12.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

# Descriptive stats for the data set, isced = 1
# NA values of income are removed

df7 = df5
df7.drop(df7[df7['income'] == 7].index, inplace = True)

df7.loc[df7['isced']==1].describe()

	age	age_rec	isced	income	wealth	Certainty_3	RiskComprehension_4	GraphLiteracy_3	Numeracy_3	Bayesianreasoning_1	Certainty_%	RiskComprehension_%	GraphLiteracy_%	Numeracy_%	Bayesianreasoning_%	TotalScore_14	TotalScore_%
count	126.000000	126.000000	126.0	126.000000	1.260000e+02	126.000000	126.000000	126.000000	126.000000	126.000000	126.000000	126.000000	126.000000	126.000000	126.000000	126.000000	126.000000
mean	33.087302	2.349206	1.0	1.888889	4.726376e+06	1.555556	1.777778	1.484127	1.460317	0.285714	51.851852	44.444444	49.470899	48.677249	28.571429	6.563492	46.882086
std	9.354374	0.541368	0.0	1.097067	1.304076e+07	0.785423	1.144649	0.952757	0.500413	0.453557	26.180758	28.616235	31.758569	16.680418	45.355737	1.745834	12.470243
min	18.000000	2.000000	1.0	1.000000	0.000000e+00	0.000000	0.000000	0.000000	1.000000	0.000000	0.000000	0.000000	0.000000	33.333333	0.000000	3.000000	21.428571
25%	27.000000	2.000000	1.0	1.000000	2.000000e+05	1.000000	1.000000	1.000000	1.000000	0.000000	33.333333	25.000000	33.333333	33.333333	0.000000	5.000000	35.714286
50%	31.000000	2.000000	1.0	2.000000	1.000000e+06	2.000000	2.000000	1.000000	1.000000	0.000000	66.666667	50.000000	33.333333	33.333333	0.000000	7.000000	50.000000
75%	37.000000	3.000000	1.0	2.000000	4.000000e+06	2.000000	3.000000	2.000000	2.000000	1.000000	66.666667	75.000000	66.666667	66.666667	100.000000	8.000000	57.142857
max	70.000000	4.000000	1.0	6.000000	1.000000e+08	3.000000	4.000000	3.000000	2.000000	1.000000	100.000000	100.000000	100.000000	66.666667	100.000000	11.000000	78.571429

df7.loc[df7['isced']==2].describe()

	age	age_rec	isced	income	wealth	Certainty_3	RiskComprehension_4	GraphLiteracy_3	Numeracy_3	Bayesianreasoning_1	Certainty_%	RiskComprehension_%	GraphLiteracy_%	Numeracy_%	Bayesianreasoning_%	TotalScore_14	TotalScore_%
count	201.000000	201.000000	201.0	201.000000	2.010000e+02	201.000000	201.000000	201.000000	201.000000	201.000000	201.000000	201.000000	201.000000	201.000000	201.000000	201.000000	201.000000
mean	27.716418	2.139303	2.0	1.706468	5.055470e+06	1.552239	1.850746	1.552239	1.412935	0.318408	51.741294	46.268657	51.741294	47.097844	31.840796	6.686567	47.761194
std	8.642579	0.387940	0.0	1.080929	1.647852e+07	0.773633	1.314387	1.033686	0.493591	0.467022	25.787755	32.859663	34.456212	16.453025	46.702190	2.268539	16.203849
min	18.000000	2.000000	2.0	1.000000	0.000000e+00	0.000000	0.000000	0.000000	1.000000	0.000000	0.000000	0.000000	0.000000	33.333333	0.000000	1.000000	7.142857
25%	22.000000	2.000000	2.0	1.000000	1.000000e+05	1.000000	1.000000	1.000000	1.000000	0.000000	33.333333	25.000000	33.333333	33.333333	0.000000	5.000000	35.714286
50%	25.000000	2.000000	2.0	1.000000	5.000000e+05	2.000000	2.000000	2.000000	1.000000	0.000000	66.666667	50.000000	66.666667	33.333333	0.000000	6.000000	42.857143
75%	31.000000	2.000000	2.0	2.000000	2.500000e+06	2.000000	3.000000	2.000000	2.000000	1.000000	66.666667	75.000000	66.666667	66.666667	100.000000	8.000000	57.142857
max	70.000000	4.000000	2.0	6.000000	1.500000e+08	3.000000	4.000000	3.000000	2.000000	1.000000	100.000000	100.000000	100.000000	66.666667	100.000000	12.000000	85.714286

df7.loc[df7['isced']==3].describe()

	age	age_rec	isced	income	wealth	Certainty_3	RiskComprehension_4	GraphLiteracy_3	Numeracy_3	Bayesianreasoning_1	Certainty_%	RiskComprehension_%	GraphLiteracy_%	Numeracy_%	Bayesianreasoning_%	TotalScore_14	TotalScore_%
count	43.000000	43.000000	43.0	43.000000	4.300000e+01	43.000000	43.000000	43.000000	43.000000	43.000000	43.000000	43.000000	43.000000	43.000000	43.000000	43.000000	43.000000
mean	22.255814	2.093023	3.0	1.837209	4.133750e+06	1.744186	1.651163	1.418605	1.488372	0.325581	58.139535	41.279070	47.286822	49.612403	32.558140	6.627907	47.342193
std	6.959382	0.293903	0.0	1.446346	1.532675e+07	0.758853	1.307191	0.879192	0.505781	0.474137	25.295108	32.679778	29.306407	16.859351	47.413732	1.826045	13.043179
min	16.000000	2.000000	3.0	1.000000	0.000000e+00	0.000000	0.000000	0.000000	1.000000	0.000000	0.000000	0.000000	0.000000	33.333333	0.000000	2.000000	14.285714
25%	18.000000	2.000000	3.0	1.000000	7.000000e+04	2.000000	1.000000	1.000000	1.000000	0.000000	66.666667	25.000000	33.333333	33.333333	0.000000	6.000000	42.857143
50%	19.000000	2.000000	3.0	1.000000	4.000000e+05	2.000000	1.000000	1.000000	1.000000	0.000000	66.666667	25.000000	33.333333	33.333333	0.000000	6.000000	42.857143
75%	22.000000	2.000000	3.0	2.000000	1.750000e+06	2.000000	3.000000	2.000000	2.000000	1.000000	66.666667	75.000000	66.666667	66.666667	100.000000	8.000000	57.142857
max	46.000000	3.000000	3.0	6.000000	1.000000e+08	3.000000	4.000000	3.000000	2.000000	1.000000	100.000000	100.000000	100.000000	66.666667	100.000000	11.000000	78.571429

# Trend line for Absolute Tota Facet Score vs income reponses sorted by isced and ORDERED by income WITH scatter

fig, ax7 = plt.subplots(figsize=(6, 6))

sns.regplot (data = df7.loc[df7['isced']==1].reset_index(drop = True), x = 'income', y = df7.loc[df7['isced']==1]['TotalScore_14'], fit_reg=True, ci=None, ax=ax7, label='ISCED = 1')
sns.regplot (data = df7.loc[df7['isced']==2].reset_index(drop = True), x = 'income', y = df7.loc[df7['isced']==2]['TotalScore_14'], fit_reg=True, ci=None, ax=ax7, label='ISCED = 2')
sns.regplot (data = df7.loc[df7['isced']==3].reset_index(drop = True), x = 'income', y = df7.loc[df7['isced']==3]['TotalScore_14'], fit_reg=True, ci=None, ax=ax7, label='ISCED = 3')

ax7.set(ylabel='Total Scores_19', xlabel='INCOME')
ax7.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

# Trend line for Absolute Tota Facet Score vs income reponses sorted by isced and ORDERED by income WITHOUT scatter

fig, ax8 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[df7['isced']==1].reset_index(drop = True), x = 'income', y = 'TotalScore_14', fit_reg=True, ci=None, ax=ax8, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[df7['isced']==2].reset_index(drop = True), x = 'income', y = 'TotalScore_14', fit_reg=True, ci=None, ax=ax8, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[df7['isced']==3].reset_index(drop = True), x = 'income', y = 'TotalScore_14', fit_reg=True, ci=None, ax=ax8, label='ISCED = 3')

ax8.set(ylabel='Total Scores_19', xlabel='INCOME')
ax8.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

# Trend line for Absolute Certainty Score vs income reponses sorted by isced and ORDERED by income WITHOUT scatter

fig, ax9 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[df7['isced']==1].reset_index(drop = True), x = 'income', y = 'Certainty_3', fit_reg=True, ci=None, ax=ax9, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[df7['isced']==2].reset_index(drop = True), x = 'income', y = 'Certainty_3', fit_reg=True, ci=None, ax=ax9, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[df7['isced']==3].reset_index(drop = True), x = 'income', y = 'Certainty_3', fit_reg=True, ci=None, ax=ax9, label='ISCED = 3')

ax9.set(ylabel='Certainty_3', xlabel='INCOME')
ax9.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

# Trend line for Absolute Risk Comprehension Score vs income reponses sorted by isced and ORDERED by income WITHOUT scatter

fig, ax10 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[df7['isced']==1].reset_index(drop = True), x = 'income', y = 'RiskComprehension_4', fit_reg=True, ci=None, ax=ax10, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[df7['isced']==2].reset_index(drop = True), x = 'income', y = 'RiskComprehension_4', fit_reg=True, ci=None, ax=ax10, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[df7['isced']==3].reset_index(drop = True), x = 'income', y = 'RiskComprehension_4', fit_reg=True, ci=None, ax=ax10, label='ISCED = 3')

ax10.set(ylabel='RiskComprehension_4', xlabel='INCOME')
ax10.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

# Trend line for Absolute Number Comprehension Score vs income reponses sorted by isced and ORDERED by income WITHOUT scatter

fig, ax11 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[df7['isced']==1].reset_index(drop = True), x = 'income', y = 'Numeracy_3', fit_reg=True, ci=None, ax=ax11, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[df7['isced']==2].reset_index(drop = True), x = 'income', y = 'Numeracy_3', fit_reg=True, ci=None, ax=ax11, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[df7['isced']==3].reset_index(drop = True), x = 'income', y = 'Numeracy_3', fit_reg=True, ci=None, ax=ax11, label='ISCED = 3')

ax11.set(ylabel='Numeracy_3', xlabel='INCOME')
ax11.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

# Trend line for Absolute Graph Comprehension Score vs income reponses sorted by isced and ORDERED by income WITHOUT scatter

fig, ax11 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[df7['isced']==1].reset_index(drop = True), x = 'income', y = 'GraphLiteracy_3', fit_reg=True, ci=None, ax=ax11, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[df7['isced']==2].reset_index(drop = True), x = 'income', y = 'GraphLiteracy_3', fit_reg=True, ci=None, ax=ax11, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[df7['isced']==3].reset_index(drop = True), x = 'income', y = 'GraphLiteracy_3', fit_reg=True, ci=None, ax=ax11, label='ISCED = 3')

ax11.set(ylabel='GraphLiteracy_3', xlabel='INCOME')
ax11.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

# Trend line for Absolute Bayesian Reasoning Score vs income reponses sorted by isced and ORDERED by income WITHOUT scatter

fig, ax12 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[df7['isced']==1].reset_index(drop = True), x = 'income', y = 'Bayesianreasoning_1', fit_reg=True, ci=None, ax=ax12, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[df7['isced']==2].reset_index(drop = True), x = 'income', y = 'Bayesianreasoning_1', fit_reg=True, ci=None, ax=ax12, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[df7['isced']==3].reset_index(drop = True), x = 'income', y = 'Bayesianreasoning_1', fit_reg=True, ci=None, ax=ax12, label='ISCED = 3')

ax12.set(ylabel='Bayesianreasoning_1', xlabel='INCOME')
ax12.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

df7

	ResponseId	age	age_rec	isced	income	wealth	Certainty_3	RiskComprehension_4	GraphLiteracy_3	Numeracy_3	Bayesianreasoning_1	Certainty_%	RiskComprehension_%	GraphLiteracy_%	Numeracy_%	Bayesianreasoning_%	TotalScore_14	TotalScore_%
0	R_4XhfVqNhg49zZXe	37	3	1	2	6000000.0	2	0	1	2	0	66.666667	0.0	33.333333	66.666667	0.0	5	35.714286
1	R_4FjN0R9SXA0s0G5	53	3	1	2	5000000.0	2	2	1	1	1	66.666667	50.0	33.333333	33.333333	100.0	7	50.000000
2	R_4Ph1UNtYnaoPSX7	25	2	1	3	400000.0	0	2	0	1	0	0.000000	50.0	0.000000	33.333333	0.0	3	21.428571
3	R_4FVQvDpbnyvqOQY	28	2	1	2	3200000.0	0	1	0	2	0	0.000000	25.0	0.000000	66.666667	0.0	3	21.428571
4	R_4PjnTbc2vIjogot	31	2	1	1	200000.0	2	1	0	1	0	66.666667	25.0	0.000000	33.333333	0.0	4	28.571429
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
365	R_4fkCPjfYl6UggHL	18	2	3	1	1000000.0	2	3	1	2	1	66.666667	75.0	33.333333	66.666667	100.0	9	64.285714
366	R_4K6WL8dnMy7ODbs	18	2	3	6	100000.0	2	0	1	2	0	66.666667	0.0	33.333333	66.666667	0.0	5	35.714286
367	R_4ANgmQQx8TBl3kc	21	2	3	1	16000000.0	0	4	2	1	1	0.000000	100.0	66.666667	33.333333	100.0	8	57.142857
368	R_4MJAzsNYYA69Y8p	27	2	3	1	0.0	2	1	1	2	1	66.666667	25.0	33.333333	66.666667	100.0	7	50.000000
369	R_4EneslabJXcCOT4	18	2	3	1	3000000.0	2	3	2	1	0	66.666667	75.0	66.666667	33.333333	0.0	8	57.142857

370 rows × 18 columns

# Since we already have a classification for Age groups in the form of age_rec, we will use that.
# We will also use median of age to see if it yields any relevant results, as instructed.

# AXES to be used = Age or age groups, ISCED, Income

# Age groups = [2,3,4]
# Age group 2 = 18 to 35 y/o
# Age group 3 = 36 to 59 y/o
# Age group 4 = 60 to 75 y/o (75 y/o, i.e, within the scope of the data we have, it can mean 60 and above also)

df7.loc[(df7['age_rec']==2)].describe()

	age	age_rec	isced	income	wealth	Certainty_3	RiskComprehension_4	GraphLiteracy_3	Numeracy_3	Bayesianreasoning_1	Certainty_%	RiskComprehension_%	GraphLiteracy_%	Numeracy_%	Bayesianreasoning_%	TotalScore_14	TotalScore_%
count	301.000000	301.0	301.000000	301.000000	3.010000e+02	301.000000	301.000000	301.000000	301.000000	301.000000	301.000000	301.000000	301.000000	301.000000	301.000000	301.000000	301.000000
mean	25.411960	2.0	1.843854	1.750831	4.944977e+06	1.581395	1.807309	1.531561	1.445183	0.312292	52.713178	45.182724	51.052049	48.172757	31.229236	6.677741	47.698149
std	4.902012	0.0	0.626262	1.122961	1.647609e+07	0.768669	1.279093	1.001582	0.497814	0.464200	25.622315	31.977333	33.386079	16.593787	46.420009	2.078253	14.844665
min	16.000000	2.0	1.000000	1.000000	0.000000e+00	0.000000	0.000000	0.000000	1.000000	0.000000	0.000000	0.000000	0.000000	33.333333	0.000000	1.000000	7.142857
25%	21.000000	2.0	1.000000	1.000000	1.000000e+05	1.000000	1.000000	1.000000	1.000000	0.000000	33.333333	25.000000	33.333333	33.333333	0.000000	5.000000	35.714286
50%	25.000000	2.0	2.000000	1.000000	5.000000e+05	2.000000	2.000000	2.000000	1.000000	0.000000	66.666667	50.000000	66.666667	33.333333	0.000000	6.000000	42.857143
75%	29.000000	2.0	2.000000	2.000000	2.500000e+06	2.000000	3.000000	2.000000	2.000000	1.000000	66.666667	75.000000	66.666667	66.666667	100.000000	8.000000	57.142857
max	35.000000	2.0	3.000000	6.000000	1.500000e+08	3.000000	4.000000	3.000000	2.000000	1.000000	100.000000	100.000000	100.000000	66.666667	100.000000	12.000000	85.714286

df7.loc[(df7['age_rec']==3)].describe()

	age	age_rec	isced	income	wealth	Certainty_3	RiskComprehension_4	GraphLiteracy_3	Numeracy_3	Bayesianreasoning_1	Certainty_%	RiskComprehension_%	GraphLiteracy_%	Numeracy_%	Bayesianreasoning_%	TotalScore_14	TotalScore_%
count	62.000000	62.0	62.000000	62.000000	6.200000e+01	62.000000	62.000000	62.000000	62.000000	62.000000	62.000000	62.000000	62.000000	62.000000	62.000000	62.000000	62.000000
mean	41.903226	3.0	1.483871	1.870968	4.204371e+06	1.564516	1.741935	1.451613	1.370968	0.306452	52.150538	43.548387	48.387097	45.698925	30.645161	6.435484	45.967742
std	5.835757	0.0	0.620469	1.108593	7.989684e+06	0.802066	1.172576	0.935238	0.487007	0.464783	26.735545	29.314394	31.174589	16.233576	46.478303	1.980404	14.145741
min	36.000000	3.0	1.000000	1.000000	1.000000e+03	0.000000	0.000000	0.000000	1.000000	0.000000	0.000000	0.000000	0.000000	33.333333	0.000000	2.000000	14.285714
25%	37.000000	3.0	1.000000	1.000000	1.500000e+05	1.000000	1.000000	1.000000	1.000000	0.000000	33.333333	25.000000	33.333333	33.333333	0.000000	5.000000	35.714286
50%	40.000000	3.0	1.000000	2.000000	1.000000e+06	2.000000	2.000000	1.000000	1.000000	0.000000	66.666667	50.000000	33.333333	33.333333	0.000000	6.500000	46.428571
75%	44.000000	3.0	2.000000	2.000000	5.000000e+06	2.000000	3.000000	2.000000	2.000000	1.000000	66.666667	75.000000	66.666667	66.666667	100.000000	8.000000	57.142857
max	59.000000	3.0	3.000000	6.000000	5.000000e+07	3.000000	4.000000	3.000000	2.000000	1.000000	100.000000	100.000000	100.000000	66.666667	100.000000	11.000000	78.571429

df7.loc[(df7['age_rec']==4)].describe()

	age	age_rec	isced	income	wealth	Certainty_3	RiskComprehension_4	GraphLiteracy_3	Numeracy_3	Bayesianreasoning_1	Certainty_%	RiskComprehension_%	GraphLiteracy_%	Numeracy_%	Bayesianreasoning_%	TotalScore_14	TotalScore_%
count	7.000000	7.0	7.000000	7.000000	7.000000e+00	7.000000	7.000000	7.000000	7.000000	7.000000	7.000000	7.000000	7.000000	7.000000	7.000000	7.000000	7.000000
mean	64.285714	4.0	1.428571	2.428571	5.759286e+06	1.428571	2.142857	1.285714	1.714286	0.142857	47.619048	53.571429	42.857143	57.142857	14.285714	6.714286	47.959184
std	4.151879	0.0	0.534522	1.718249	4.039660e+06	0.975900	1.069045	0.951190	0.487950	0.377964	32.530002	26.726124	31.706324	16.265001	37.796447	1.496026	10.685903
min	61.000000	4.0	1.000000	1.000000	3.500000e+05	0.000000	0.000000	0.000000	1.000000	0.000000	0.000000	0.000000	0.000000	33.333333	0.000000	5.000000	35.714286
25%	61.000000	4.0	1.000000	1.500000	2.750000e+06	1.000000	2.000000	0.500000	1.500000	0.000000	33.333333	50.000000	16.666667	50.000000	0.000000	5.500000	39.285714
50%	62.000000	4.0	1.000000	2.000000	7.000000e+06	2.000000	2.000000	2.000000	2.000000	0.000000	66.666667	50.000000	66.666667	66.666667	0.000000	7.000000	50.000000
75%	67.500000	4.0	2.000000	2.500000	8.732500e+06	2.000000	3.000000	2.000000	2.000000	0.000000	66.666667	75.000000	66.666667	66.666667	0.000000	7.500000	53.571429
max	70.000000	4.0	2.000000	6.000000	1.000000e+07	2.000000	3.000000	2.000000	2.000000	1.000000	66.666667	75.000000	66.666667	66.666667	100.000000	9.000000	64.285714

fig, ax1 = plt.subplots(figsize=(6, 6))

sns.regplot (data = df7.loc[(df7['age_rec']==2) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==2) & (df7['isced']==1)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 1')
sns.regplot (data = df7.loc[(df7['age_rec']==2) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==2) & (df7['isced']==2)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 2')
sns.regplot (data = df7.loc[(df7['age_rec']==2) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==2) & (df7['isced']==3)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 3')

ax1.set(ylabel='Total Scores_14', xlabel='INCOME 18 to 35')
ax1.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

fig, ax2 = plt.subplots(figsize=(6, 6))

sns.regplot (data = df7.loc[(df7['age_rec']==3) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==3) & (df7['isced']==1)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 1')
sns.regplot (data = df7.loc[(df7['age_rec']==3) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==3) & (df7['isced']==2)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 2')
sns.regplot (data = df7.loc[(df7['age_rec']==3) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==3) & (df7['isced']==3)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 3')

ax2.set(ylabel='Total Scores_14', xlabel='INCOME 36 to 59')
ax2.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

fig, ax3 = plt.subplots(figsize=(6, 6))

sns.regplot (data = df7.loc[(df7['age_rec']==4) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==4) & (df7['isced']==1)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax3, label='ISCED = 1')
sns.regplot (data = df7.loc[(df7['age_rec']==4) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==4) & (df7['isced']==2)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax3, label='ISCED = 2')
sns.regplot (data = df7.loc[(df7['age_rec']==4) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==4) & (df7['isced']==3)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax3, label='ISCED = 3')

ax3.set(ylabel='Total Scores_14', xlabel='INCOME 60 to 75')
ax3.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

fig, ax1 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==2) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==2) & (df7['isced']==1)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==2) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==2) & (df7['isced']==2)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==2) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==2) & (df7['isced']==3)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 3')

ax1.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ df7['age_rec']==2  ].groupby(['income']).mean(numeric_only=True)['TotalScore_14'], yerr = df7.loc[ (df7['age_rec']==2) ].groupby(['income'])['TotalScore_14'].sem(), fmt='o', color = lighten_color('gray',0.5))
ax1.set(ylabel='Total Scores_14', xlabel='INCOME 18 to 35',  yticks = np.arange(5, 16 , 1))
ax1.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

fig, ax2 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==3) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==3) & (df7['isced']==1)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==3) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==3) & (df7['isced']==2)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==3) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==3) & (df7['isced']==3)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 3')

ax2.set(ylabel='Total Scores_14', xlabel='INCOME 36 to 59',  yticks = np.arange(5, 16 , 1))
ax2.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
ax2.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ df7['age_rec']==3  ].groupby(['income']).mean(numeric_only=True)['TotalScore_14'], yerr = df7.loc[ (df7['age_rec']==3) ].groupby(['income'])['TotalScore_14'].sem(), fmt='o', color = lighten_color('gray',0.5))
plt.show()

fig, ax3 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==4) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==4) & (df7['isced']==1)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax3, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==4) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==4) & (df7['isced']==2)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax3, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==4) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==4) & (df7['isced']==3)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax3, label='ISCED = 3')

ax3.set(ylabel='Total Scores_14', xlabel='INCOME 60 to 75',  yticks = np.arange(5, 16 , 1))
ax3.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
ax3.errorbar(x = df7.loc[ df7['age_rec']== 4 ]['income'].unique(), y = df7.loc[ df7['age_rec']==4  ].groupby(['income']).mean(numeric_only=True)['TotalScore_14'], yerr = df7.loc[ (df7['age_rec']==4) ].groupby(['income'])['TotalScore_14'].sem(), fmt='o', color = lighten_color('gray',0.5))
plt.show()

fig, ax1 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==2) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==2) & (df7['isced']==1)]['Certainty_3'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==2) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==2) & (df7['isced']==2)]['Certainty_3'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==2) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==2) & (df7['isced']==3)]['Certainty_3'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 3')

ax1.set(ylabel='Certainty_3', xlabel='INCOME 18 to 35',  yticks = np.arange(0, 6 , 1))
ax1.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
ax1.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ df7['age_rec']==2  ].groupby(['income'])['Certainty_3'].mean(), yerr = df7.loc[ (df7['age_rec']==2) ].groupby(['income'])['Certainty_3'].sem(), fmt='o', color = lighten_color('gray',0.5))
plt.show()

fig, ax2 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==3) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==3) & (df7['isced']==1)]['Certainty_3'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==3) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==3) & (df7['isced']==2)]['Certainty_3'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==3) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==3) & (df7['isced']==3)]['Certainty_3'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 3')

ax2.set(ylabel='Certainty_3', xlabel='INCOME 36 to 59',  yticks = np.arange(0, 6 , 1))
ax2.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
ax2.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ df7['age_rec']==3  ].groupby(['income'])['Certainty_3'].mean(), yerr = df7.loc[ (df7['age_rec']==3) ].groupby(['income'])['Certainty_3'].sem(), fmt='o', color = lighten_color('gray',0.5))
plt.show()

fig, ax3 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==4) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==4) & (df7['isced']==1)]['Certainty_3'], fit_reg=True, ci=None, ax=ax3, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==4) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==4) & (df7['isced']==2)]['Certainty_3'], fit_reg=True, ci=None, ax=ax3, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==4) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==4) & (df7['isced']==3)]['Certainty_3'], fit_reg=True, ci=None, ax=ax3, label='ISCED = 3')

ax3.set(ylabel='Certainty_3', xlabel='INCOME 60 to 75',  yticks = np.arange(0, 6 , 1))
ax3.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
ax3.errorbar(x = df7.loc[ df7['age_rec']== 4 ]['income'].unique(), y = df7.loc[ df7['age_rec']==4  ].groupby(['income'])['Certainty_3'].mean(), yerr = df7.loc[ (df7['age_rec']==4) ].groupby(['income'])['Certainty_3'].sem(), fmt='o', color = lighten_color('gray',0.5))
plt.show()

fig, ax1 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==2) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==2) & (df7['isced']==1)]['RiskComprehension_4'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==2) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==2) & (df7['isced']==2)]['RiskComprehension_4'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==2) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==2) & (df7['isced']==3)]['RiskComprehension_4'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 3')

ax1.set(ylabel='RiskComprehension_4', xlabel='INCOME 18 to 35',  yticks = np.arange(0, 3 , 1))
ax1.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
ax1.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ df7['age_rec']==2  ].groupby(['income'])['RiskComprehension_4'].mean(), yerr = df7.loc[ (df7['age_rec']==2) ].groupby(['income'])['RiskComprehension_4'].sem(), fmt='o', color = lighten_color('gray',0.5))
plt.show()

fig, ax2 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==3) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==3) & (df7['isced']==1)]['RiskComprehension_4'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==3) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==3) & (df7['isced']==2)]['RiskComprehension_4'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==3) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==3) & (df7['isced']==3)]['RiskComprehension_4'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 3')

ax2.set(ylabel='RiskComprehension_4', xlabel='INCOME 36 to 59',  yticks = np.arange(0, 3 , 1))
ax2.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
ax2.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ df7['age_rec']==3  ].groupby(['income'])['RiskComprehension_4'].mean(), yerr = df7.loc[ (df7['age_rec']==3) ].groupby(['income'])['RiskComprehension_4'].sem(), fmt='o', color = lighten_color('gray',0.5))
plt.show()

fig, ax3 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==4) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==4) & (df7['isced']==1)]['RiskComprehension_4'], fit_reg=True, ci=None, ax=ax3, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==4) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==4) & (df7['isced']==2)]['RiskComprehension_4'], fit_reg=True, ci=None, ax=ax3, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==4) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==4) & (df7['isced']==3)]['RiskComprehension_4'], fit_reg=True, ci=None, ax=ax3, label='ISCED = 3')

ax3.set(ylabel='RiskComprehension_4', xlabel='INCOME 60 to 75',  yticks = np.arange(0, 3 , 1))
ax3.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
ax3.errorbar(x = df7.loc[ df7['age_rec']== 4 ]['income'].unique(), y = df7.loc[ df7['age_rec']==4  ].groupby(['income'])['RiskComprehension_4'].mean(), yerr = df7.loc[ (df7['age_rec']==4) ].groupby(['income'])['RiskComprehension_4'].sem(), fmt='o', color = lighten_color('gray',0.5))
plt.show()

fig, ax1 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==2) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==2) & (df7['isced']==1)]['Numeracy_3'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==2) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==2) & (df7['isced']==2)]['Numeracy_3'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==2) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==2) & (df7['isced']==3)]['Numeracy_3'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 3')

ax1.set(ylabel='Numeracy_3', xlabel='INCOME 18 to 35',  yticks = np.arange(0, 6 , 1))
ax1.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
ax1.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ df7['age_rec']==2  ].groupby(['income'])['Numeracy_3'].mean(), yerr = df7.loc[ (df7['age_rec']==2) ].groupby(['income'])['Numeracy_3'].sem(), fmt='o', color = lighten_color('gray',0.5))
plt.show()

fig, ax2 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==3) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==3) & (df7['isced']==1)]['Numeracy_3'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==3) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==3) & (df7['isced']==2)]['Numeracy_3'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==3) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==3) & (df7['isced']==3)]['Numeracy_3'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 3')

ax2.set(ylabel='Numeracy_3', xlabel='INCOME 36 to 59',  yticks = np.arange(0, 6 , 1))
ax2.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
ax2.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ df7['age_rec']==3  ].groupby(['income'])['Numeracy_3'].mean(), yerr = df7.loc[ (df7['age_rec']==3) ].groupby(['income'])['Numeracy_3'].sem(), fmt='o', color = lighten_color('gray',0.5))
plt.show()

fig, ax3 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==4) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==4) & (df7['isced']==1)]['Numeracy_3'], fit_reg=True, ci=None, ax=ax3, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==4) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==4) & (df7['isced']==2)]['Numeracy_3'], fit_reg=True, ci=None, ax=ax3, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==4) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==4) & (df7['isced']==3)]['Numeracy_3'], fit_reg=True, ci=None, ax=ax3, label='ISCED = 3')

ax3.set(ylabel='Numeracy_3', xlabel='INCOME 60 to 75',  yticks = np.arange(0, 6 , 1))
ax3.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
ax3.errorbar(x = df7.loc[ df7['age_rec']== 4 ]['income'].unique(), y = df7.loc[ df7['age_rec']==4  ].groupby(['income'])['Numeracy_3'].mean(), yerr = df7.loc[ (df7['age_rec']==4) ].groupby(['income'])['Numeracy_3'].sem(), fmt='o', color = lighten_color('gray',0.5))
plt.show()

fig, ax1 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==2) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==2) & (df7['isced']==1)]['GraphLiteracy_3'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==2) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==2) & (df7['isced']==2)]['GraphLiteracy_3'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==2) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==2) & (df7['isced']==3)]['GraphLiteracy_3'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 3')

ax1.set(ylabel='GraphLiteracy_3', xlabel='INCOME 18 to 35',  yticks = np.arange(0, 4 , 1))
ax1.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
ax1.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ df7['age_rec']==2  ].groupby(['income'])['GraphLiteracy_3'].mean(), yerr = df7.loc[ (df7['age_rec']==2) ].groupby(['income'])['GraphLiteracy_3'].sem(), fmt='o', color = lighten_color('gray',0.5))
plt.show()

fig, ax2 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==3) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==3) & (df7['isced']==1)]['GraphLiteracy_3'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==3) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==3) & (df7['isced']==2)]['GraphLiteracy_3'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==3) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==3) & (df7['isced']==3)]['GraphLiteracy_3'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 3')

ax2.set(ylabel='GraphLiteracy_3', xlabel='INCOME 36 to 59',  yticks = np.arange(0, 4 , 1))
ax2.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
ax2.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ df7['age_rec']==3  ].groupby(['income'])['GraphLiteracy_3'].mean(), yerr = df7.loc[ (df7['age_rec']==3) ].groupby(['income'])['GraphLiteracy_3'].sem(), fmt='o', color = lighten_color('gray',0.5))
plt.show()

fig, ax3 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==4) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==4) & (df7['isced']==1)]['GraphLiteracy_3'], fit_reg=True, ci=None, ax=ax3, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==4) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==4) & (df7['isced']==2)]['GraphLiteracy_3'], fit_reg=True, ci=None, ax=ax3, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==4) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==4) & (df7['isced']==3)]['GraphLiteracy_3'], fit_reg=True, ci=None, ax=ax3, label='ISCED = 3')

ax3.set(ylabel='GraphLiteracy_3', xlabel='INCOME 60 to 75',  yticks = np.arange(0, 4 , 1))
ax3.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
ax3.errorbar(x = df7.loc[ df7['age_rec']== 4 ]['income'].unique(), y = df7.loc[ df7['age_rec']==4  ].groupby(['income'])['GraphLiteracy_3'].mean(), yerr = df7.loc[ (df7['age_rec']==4) ].groupby(['income'])['GraphLiteracy_3'].sem(), fmt='o', color = lighten_color('gray',0.5))
plt.show()

fig, ax1 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==2) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==2) & (df7['isced']==1)]['Bayesianreasoning_1'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==2) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==2) & (df7['isced']==2)]['Bayesianreasoning_1'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==2) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==2) & (df7['isced']==3)]['Bayesianreasoning_1'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 3')

ax1.set(ylabel='Bayesianreasoning_1', xlabel='INCOME 18 to 35',  yticks = np.arange(0, 5 , 1))
ax1.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
ax1.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ df7['age_rec']==2  ].groupby(['income'])['Bayesianreasoning_1'].mean(), yerr = df7.loc[ (df7['age_rec']==2) ].groupby(['income'])['Bayesianreasoning_1'].sem(), fmt='o', color = lighten_color('gray',0.5))
plt.show()

fig, ax2 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==3) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==3) & (df7['isced']==1)]['Bayesianreasoning_1'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==3) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==3) & (df7['isced']==2)]['Bayesianreasoning_1'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==3) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==3) & (df7['isced']==3)]['Bayesianreasoning_1'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 3')

ax2.set(ylabel='Bayesianreasoning_1', xlabel='INCOME 36 to 59',  yticks = np.arange(0, 5 , 1))
ax2.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
ax2.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ df7['age_rec']==3  ].groupby(['income'])['Bayesianreasoning_1'].mean(), yerr = df7.loc[ (df7['age_rec']==3) ].groupby(['income'])['Bayesianreasoning_1'].sem(), fmt='o', color = lighten_color('gray',0.5))
plt.show()

fig, ax3 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==4) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==4) & (df7['isced']==1)]['Bayesianreasoning_1'], fit_reg=True, ci=None, ax=ax3, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==4) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==4) & (df7['isced']==2)]['Bayesianreasoning_1'], fit_reg=True, ci=None, ax=ax3, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age_rec']==4) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age_rec']==4) & (df7['isced']==3)]['Bayesianreasoning_1'], fit_reg=True, ci=None, ax=ax3, label='ISCED = 3')

ax3.set(ylabel='Bayesianreasoning_1', xlabel='INCOME 60 to 75',  yticks = np.arange(0, 5 , 1))
ax3.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
ax3.errorbar(x = df7.loc[ df7['age_rec']== 4 ]['income'].unique(), y = df7.loc[ df7['age_rec']==4  ].groupby(['income'])['Bayesianreasoning_1'].mean(), yerr = df7.loc[ (df7['age_rec']==4) ].groupby(['income'])['Bayesianreasoning_1'].sem(), fmt='o', color = lighten_color('gray',0.5))
plt.show()

# Age median method
# Age_median

age_m = df1["age"].median()
age_m

27.0

df7["income"].value_counts()

income
1    199
2    106
3     35
4     15
6      9
5      6
Name: count, dtype: int64

df7.loc[(df7['age'] < age_m)].describe()

	age	age_rec	isced	income	wealth	Certainty_3	RiskComprehension_4	GraphLiteracy_3	Numeracy_3	Bayesianreasoning_1	Certainty_%	RiskComprehension_%	GraphLiteracy_%	Numeracy_%	Bayesianreasoning_%	TotalScore_14	TotalScore_%
count	175.000000	175.0	175.000000	175.000000	1.750000e+02	175.000000	175.000000	175.000000	175.000000	175.000000	175.000000	175.000000	175.000000	175.000000	175.000000	175.000000	175.000000
mean	21.868571	2.0	2.034286	1.720000	4.902680e+06	1.662857	1.897143	1.497143	1.417143	0.320000	55.428571	47.428571	49.904762	47.238095	32.000000	6.794286	48.530612
std	2.586411	0.0	0.614921	1.196931	1.658430e+07	0.731510	1.322112	0.964152	0.494502	0.467815	24.383663	33.052801	32.138407	16.483393	46.781468	2.134277	15.244838
min	16.000000	2.0	1.000000	1.000000	0.000000e+00	0.000000	0.000000	0.000000	1.000000	0.000000	0.000000	0.000000	0.000000	33.333333	0.000000	1.000000	7.142857
25%	20.000000	2.0	2.000000	1.000000	1.000000e+05	1.000000	1.000000	1.000000	1.000000	0.000000	33.333333	25.000000	33.333333	33.333333	0.000000	5.000000	35.714286
50%	22.000000	2.0	2.000000	1.000000	5.000000e+05	2.000000	2.000000	2.000000	1.000000	0.000000	66.666667	50.000000	66.666667	33.333333	0.000000	6.000000	42.857143
75%	24.000000	2.0	2.000000	2.000000	1.800000e+06	2.000000	3.000000	2.000000	2.000000	1.000000	66.666667	75.000000	66.666667	66.666667	100.000000	8.000000	57.142857
max	26.000000	2.0	3.000000	6.000000	1.500000e+08	3.000000	4.000000	3.000000	2.000000	1.000000	100.000000	100.000000	100.000000	66.666667	100.000000	11.000000	78.571429

df7.loc[(df7['age'] > age_m)].describe()

	age	age_rec	isced	income	wealth	Certainty_3	RiskComprehension_4	GraphLiteracy_3	Numeracy_3	Bayesianreasoning_1	Certainty_%	RiskComprehension_%	GraphLiteracy_%	Numeracy_%	Bayesianreasoning_%	TotalScore_14	TotalScore_%
count	177.000000	177.000000	177.000000	177.000000	1.770000e+02	177.000000	177.000000	177.000000	177.000000	177.000000	177.000000	177.000000	177.000000	177.000000	177.000000	177.000000	177.000000
mean	36.067797	2.429379	1.525424	1.847458	3.937876e+06	1.497175	1.711864	1.491525	1.468927	0.293785	49.905838	42.796610	49.717514	48.964218	29.378531	6.463277	46.166263
std	8.701223	0.570920	0.564759	1.078934	9.981187e+06	0.798609	1.197310	1.023134	0.500449	0.456787	26.620294	29.932754	34.104460	16.681640	45.678697	1.994326	14.245187
min	28.000000	2.000000	1.000000	1.000000	0.000000e+00	0.000000	0.000000	0.000000	1.000000	0.000000	0.000000	0.000000	0.000000	33.333333	0.000000	2.000000	14.285714
25%	30.000000	2.000000	1.000000	1.000000	1.650000e+05	1.000000	1.000000	1.000000	1.000000	0.000000	33.333333	25.000000	33.333333	33.333333	0.000000	5.000000	35.714286
50%	33.000000	2.000000	1.000000	2.000000	6.000000e+05	2.000000	2.000000	1.000000	1.000000	0.000000	66.666667	50.000000	33.333333	33.333333	0.000000	6.000000	42.857143
75%	39.000000	3.000000	2.000000	2.000000	4.500000e+06	2.000000	3.000000	2.000000	2.000000	1.000000	66.666667	75.000000	66.666667	66.666667	100.000000	8.000000	57.142857
max	70.000000	4.000000	3.000000	6.000000	1.000000e+08	3.000000	4.000000	3.000000	2.000000	1.000000	100.000000	100.000000	100.000000	66.666667	100.000000	12.000000	85.714286

fig, ax1 = plt.subplots(figsize=(6, 6))

sns.regplot (data = df7.loc[(df7['age'] < age_m) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] < age_m) & (df7['isced']==1)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 1')
sns.regplot (data = df7.loc[(df7['age'] < age_m) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] < age_m) & (df7['isced']==2)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 2')
sns.regplot (data = df7.loc[(df7['age'] < age_m) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] < age_m) & (df7['isced']==3)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 3')

ax1.set(ylabel='Total Scores_19', xlabel='INCOME for < median age')
ax1.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

fig, ax2 = plt.subplots(figsize=(6, 6))

sns.regplot (data = df7.loc[(df7['age'] > age_m) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] > age_m) & (df7['isced']==1)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 1')
sns.regplot (data = df7.loc[(df7['age'] > age_m) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] > age_m) & (df7['isced']==2)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 2')
sns.regplot (data = df7.loc[(df7['age'] > age_m) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] > age_m) & (df7['isced']==3)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 3')

ax2.set(ylabel='Total Scores_19', xlabel='INCOME for > median age')
ax2.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

fig, ax1 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age'] < age_m) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] < age_m) & (df7['isced']==1)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] < age_m) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] < age_m) & (df7['isced']==2)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] < age_m) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] < age_m) & (df7['isced']==3)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 3')

ax1.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ (df7['age'] < age_m)  ].groupby(['income'])['TotalScore_14'].mean(), yerr = df7.loc[ (df7['age'] < age_m) ].groupby(['income'])['TotalScore_14'].sem(), fmt='o', color = lighten_color('gray',0.5))
ax1.set(ylabel='Total Scores_19', xlabel='INCOME for < median age',  yticks = np.arange(5, 16 , 1))
ax1.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

fig, ax2 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age'] > age_m) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] > age_m) & (df7['isced']==1)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] > age_m) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] > age_m) & (df7['isced']==2)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] > age_m) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] > age_m) & (df7['isced']==3)]['TotalScore_14'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 3')

ax2.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ (df7['age'] > age_m)  ].groupby(['income'])['TotalScore_14'].mean(), yerr = df7.loc[ (df7['age'] > age_m) ].groupby(['income'])['TotalScore_14'].sem(), fmt='o', color = lighten_color('gray',0.5))
ax2.set(ylabel='Total Scores_19', xlabel='INCOME for > median age',  yticks = np.arange(5, 16 , 1))
ax2.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

fig, ax1 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age'] < age_m) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] < age_m) & (df7['isced']==1)]['Certainty_3'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] < age_m) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] < age_m) & (df7['isced']==2)]['Certainty_3'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] < age_m) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] < age_m) & (df7['isced']==3)]['Certainty_3'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 3')

ax1.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ (df7['age'] < age_m)  ].groupby(['income'])['Certainty_3'].mean(), yerr = df7.loc[ (df7['age'] < age_m) ].groupby(['income'])['Certainty_3'].sem(), fmt='o', color = lighten_color('gray',0.5))
ax1.set(ylabel='Certainty_3', xlabel='INCOME for < median age',  yticks = np.arange(0, 6 , 1))
ax1.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

fig, ax2 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age'] > age_m) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] > age_m) & (df7['isced']==1)]['Certainty_3'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] > age_m) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] > age_m) & (df7['isced']==2)]['Certainty_3'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] > age_m) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] > age_m) & (df7['isced']==3)]['Certainty_3'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 3')

ax2.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ (df7['age'] > age_m)  ].groupby(['income'])['Certainty_3'].mean(), yerr = df7.loc[ (df7['age'] > age_m) ].groupby(['income'])['Certainty_3'].sem(), fmt='o', color = lighten_color('gray',0.5))
ax2.set(ylabel='Certainty_3', xlabel='INCOME for > median age')
ax2.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

fig, ax1 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age'] < age_m) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] < age_m) & (df7['isced']==1)]['RiskComprehension_4'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] < age_m) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] < age_m) & (df7['isced']==2)]['RiskComprehension_4'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] < age_m) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] < age_m) & (df7['isced']==3)]['RiskComprehension_4'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 3')

ax1.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ (df7['age'] < age_m)  ].groupby(['income'])['RiskComprehension_4'].mean(), yerr = df7.loc[ (df7['age'] < age_m) ].groupby(['income'])['RiskComprehension_4'].sem(), fmt='o', color = lighten_color('gray',0.5))
ax1.set(ylabel='RiskComprehension_4', xlabel='INCOME for < median age',  yticks = np.arange(0, 3 , 1))
ax1.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

fig, ax2 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age'] > age_m) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] > age_m) & (df7['isced']==1)]['RiskComprehension_4'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] > age_m) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] > age_m) & (df7['isced']==2)]['RiskComprehension_4'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] > age_m) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] > age_m) & (df7['isced']==3)]['RiskComprehension_4'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 3')

ax2.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ (df7['age'] > age_m)  ].groupby(['income'])['RiskComprehension_4'].mean(), yerr = df7.loc[ (df7['age'] > age_m) ].groupby(['income'])['RiskComprehension_4'].sem(), fmt='o', color = lighten_color('gray',0.5))
ax2.set(ylabel='RiskComprehension_4', xlabel='INCOME for > median age',  yticks = np.arange(0, 3 , 1))
ax2.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

fig, ax1 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age'] < age_m) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] < age_m) & (df7['isced']==1)]['Numeracy_3'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] < age_m) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] < age_m) & (df7['isced']==2)]['Numeracy_3'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] < age_m) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] < age_m) & (df7['isced']==3)]['Numeracy_3'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 3')

ax1.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ (df7['age'] < age_m)  ].groupby(['income'])['Numeracy_3'].mean(), yerr = df7.loc[ (df7['age'] < age_m) ].groupby(['income'])['Numeracy_3'].sem(), fmt='o', color = lighten_color('gray',0.5))
ax1.set(ylabel='Numeracy_3', xlabel='INCOME for < median age',  yticks = np.arange(0, 6 , 1))
ax1.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

fig, ax2 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age'] > age_m) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] > age_m) & (df7['isced']==1)]['Numeracy_3'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] > age_m) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] > age_m) & (df7['isced']==2)]['Numeracy_3'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] > age_m) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] > age_m) & (df7['isced']==3)]['Numeracy_3'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 3')

ax2.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ (df7['age'] > age_m)  ].groupby(['income'])['Numeracy_3'].mean(), yerr = df7.loc[ (df7['age'] > age_m) ].groupby(['income'])['Numeracy_3'].sem(), fmt='o', color = lighten_color('gray',0.5))
ax2.set(ylabel='Numeracy_3', xlabel='INCOME for > median age',  yticks = np.arange(0, 6 , 1))
ax2.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

fig, ax1 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age'] < age_m) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] < age_m) & (df7['isced']==1)]['GraphLiteracy_3'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] < age_m) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] < age_m) & (df7['isced']==2)]['GraphLiteracy_3'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] < age_m) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] < age_m) & (df7['isced']==3)]['GraphLiteracy_3'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 3')

ax1.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ (df7['age'] < age_m)  ].groupby(['income'])['GraphLiteracy_3'].mean(), yerr = df7.loc[ (df7['age'] < age_m) ].groupby(['income'])['GraphLiteracy_3'].sem(), fmt='o', color = lighten_color('gray',0.5))
ax1.set(ylabel='GraphLiteracy_3', xlabel='INCOME for < median age',  yticks = np.arange(0, 4 , 1))
ax1.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

fig, ax2 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age'] > age_m) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] > age_m) & (df7['isced']==1)]['GraphLiteracy_3'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] > age_m) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] > age_m) & (df7['isced']==2)]['GraphLiteracy_3'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] > age_m) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] > age_m) & (df7['isced']==3)]['GraphLiteracy_3'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 3')

ax2.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ (df7['age'] > age_m)  ].groupby(['income'])['GraphLiteracy_3'].mean(), yerr = df7.loc[ (df7['age'] > age_m) ].groupby(['income'])['GraphLiteracy_3'].sem(), fmt='o', color = lighten_color('gray',0.5))
ax2.set(ylabel='GraphLiteracy_3', xlabel='INCOME for > median age',  yticks = np.arange(0, 4 , 1))
ax2.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

fig, ax1 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age'] < age_m) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] < age_m) & (df7['isced']==1)]['Bayesianreasoning_1'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] < age_m) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] < age_m) & (df7['isced']==2)]['Bayesianreasoning_1'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] < age_m) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] < age_m) & (df7['isced']==3)]['Bayesianreasoning_1'], fit_reg=True, ci=None, ax=ax1, label='ISCED = 3')

ax1.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ (df7['age'] < age_m)  ].groupby(['income'])['Bayesianreasoning_1'].mean(), yerr = df7.loc[ (df7['age'] < age_m) ].groupby(['income'])['Bayesianreasoning_1'].sem(), fmt='o', color = lighten_color('gray',0.5))
ax1.set(ylabel='Bayesianreasoning_1', xlabel='INCOME for < median age',  yticks = np.arange(0, 5 , 1))
ax1.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()

fig, ax2 = plt.subplots(figsize=(6, 6))

sns.regplot (scatter = False, data = df7.loc[(df7['age'] > age_m) & (df7['isced']==1)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] > age_m) & (df7['isced']==1)]['Bayesianreasoning_1'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 1')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] > age_m) & (df7['isced']==2)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] > age_m) & (df7['isced']==2)]['Bayesianreasoning_1'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 2')
sns.regplot (scatter = False, data = df7.loc[(df7['age'] > age_m) & (df7['isced']==3)].reset_index(drop = True), x = 'income', y = df7.loc[(df7['age'] > age_m) & (df7['isced']==3)]['Bayesianreasoning_1'], fit_reg=True, ci=None, ax=ax2, label='ISCED = 3')

ax2.errorbar(x = [1,2,3,4,5,6], y = df7.loc[ (df7['age'] > age_m)  ].groupby(['income'])['Bayesianreasoning_1'].mean(), yerr = df7.loc[ (df7['age'] > age_m) ].groupby(['income'])['Bayesianreasoning_1'].sem(), fmt='o', color = lighten_color('gray',0.5))
ax2.set(ylabel='Bayesianreasoning_1', xlabel='INCOME for > median age',  yticks = np.arange(0, 5 , 1))
ax2.legend(loc='center left',bbox_to_anchor=(1.0, 0.5))
plt.show()